INDIVIDUAL DIFFERENCES AND ADOLESCENT 
PSYCHOSOCIAL DEVELOPMENT 
 
 
 
 
 
 
 
 
by 
LESA K. ELLIS 
 
 
 
 
A DISSERTATION 
 
Presented to the Department of Psychology 
and the Graduate School of the University of Oregon 
in partial fulfillment of the requirements 
for the degree of 
Doctor of Philosophy 
 
August 2002 
 ii
 
 
 
“Individual Differences and Adolescent Psychosocial Development,” a 
dissertation prepared by Lesa K. Ellis in partial fulfillment of the requirements for 
the Doctor of Philosophy degree in the Department of Psychology.  This 
dissertation has been approved and accepted by:  
 
 
 
 
 
____________________________________________________________ 
Dr. Mary K. Rothbart, Chair of the Examining Committee 
 
 
 
________________________________________ 
Date 
 
 
 
Committee in charge:   Dr. Mary K. Rothbart, Chair 
    Dr. Kirby Deater-Deckard 
    Dr. Thomas J. Dishion 
    Dr. Roland H. Good 
 
 
Accepted by 
 
 
____________________________________________________________ 
Vice-Provost and Dean of the Graduate School 
 
 
 iii
 
CURRICULUM VITA 
NAME OF AUTHOR:  Lesa K. Ellis 
 
PLACE OF BIRTH:  American Fork, Utah 
 
DATE OF BIRTH: April 30, 1959 
 
GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED: 
 
 University of Oregon 
 Westminster College 
 Salt Lake Community College 
  
 
DEGREES AWARDED: 
 
 Doctor of Philosophy in Psychology, 2002, University of Oregon 
 Master of Science in Psychology, 1999, University of Oregon 
 Bachelor of Science in Psychology, 1998, Westminster College  
 
AREAS OF SPECIAL INTEREST: 
 
 Adolescent Development 
 Affective Neuroscience 
 
PROFESSIONAL EXPERIENCE: 
 
 Instructor, Adolescent Development, Department of Psychology, University of 
Oregon, 2001-2002 
 Instructor, Applied Statistics, Department of Psychology, University of Oregon, 
2002 
 Research Assistant, Attention and Temperament Laboratory, Department of 
Psychology, University of Oregon, 1998-2002 
  
AWARDS AND HONORS: 
 
 Beverly Fagot Dissertation Fellowship, University of Oregon, 2002 
 NIMH Emotion Research Pre-doctoral Fellowship, University of Oregon, 1999-
2001 
 Eccles foundation Student-Faculty Development Grant, Westminster College, 
1996 
 
 iv
 
PUBLICATIONS 
 
 Ellis, L. K., & Rothbart, M. K. (2002).  Revision of the Early Adolescent 
Temperament Questionnaire.  Manuscript in preparation. 
 
 Putnam, S. P., Ellis, L. K., & Rothbart, M. K. (2001).  The structure of 
temperament from infancy through adolescence.  In A. Eliasz & A. Angleitner 
(Eds.) Advances in Research on Temperament (pp. 185-162).  Germany: Pabst 
Scientist Publisher. 
 
 Rothbart, M. K., Ellis, L. K., & Posner, M. I.  (2002).  Developing mechanisms of 
temperamental effortful control.  Manuscript submitted for publication. 
 
 v
 
 
ACKNOWLEDGMENTS 
 
The author expresses sincere gratitude to Dr. Mary Rothbart for her excellent 
guidance and mentorship over the years.  In addition, special thanks are due to the other 
committee members for providing the expertise and resources to make this project 
possible.  I also wish to thank Dr. Michael I. Posner who, while not an official committee 
member, was consistently generous with his time and knowledge.  This project was 
supported in part by a National Research Service Predoctoral Training Grant, Number 5 
T32 MH18935-10, from the National Institute of Mental Health, by a National Institute 
of Mental Health grant, Number 43361, to Dr. Mary K. Rothbart at the University of 
Oregon, and by a grant from the National Institute on Drug Abuse at the National 
Institutes of Health, Number DA 07031, to Dr. Thomas J. Dishion at the University of 
Oregon. 
 vi
 
DEDICATION 
 
 This project is dedicated to my family, without whom my education would still be 
nothing more than a distant dream. 
 
 vii
 
TABLE OF CONTENTS 
Chapter                                                                                                                 Page 
I.  INTRODUCTION ...........................................................................................15 
Temperament, Attention, and Puberty as Individual Differences...........17 
Assessing Puberty ...............................................................................19 
Adolescent Emotional Reactivity and Psychosocial Development .......24 
Negative Affectivity ...........................................................................24 
Puberty and Negative Affect ..............................................................28 
Approach..............................................................................................33 
Affiliativeness ......................................................................................37 
Adolescent Self-Regulation and Psychosocial Development ................40 
Effortful Control ..................................................................................40 
Parenting, Peers, and Psychosocial Outcomes.........................................42 
Executive Attentional Systems.................................................................43 
Preliminary Studies...................................................................................46 
The Current Studies ..................................................................................50 
II.  STUDY I ........................................................................................................53 
Introduction ..............................................................................................53 
Participants................................................................................................53 
Measures ...................................................................................................54 
Temperament Questionnaire ...............................................................54 
Pubertal Status ............................................................................58 
General Procedure .......................................................................60 
Gender Differences .....................................................................60 
Results .......................................................................................................60 
Descriptive and Psychometric Properties of the EATQ-R ................60 
Agreement between Reporters ...................................................60 
Gender Differences .....................................................................64 
Correlations between Temperament Scales and 
Socio-Emotional Scales ........................................................64 
Body Changes Questionnaire .............................................................65 
Age and Gender Differences ......................................................66 
Agreement between Raters..........................................................67 
Temperament and Puberty ...................................................................69 
Puberty Categories ......................................................................69 
Menarcheal Status .......................................................................73 
Pubertal Status Score, Age, and Temperament .........................75 
Effortful Control ...................................................................80 
Affiliativeness .......................................................................81 
 
 viii
Negative Affectivity .............................................................81 
Surgency ...............................................................................82 
Temperament, Puberty, and Socio-Emotional Behaviors.......................82 
Discussion ................................................................................................85 
Psychometric Properties of the EATQ-R............................................85 
Temperament, Puberty, and Age.........................................................88 
Temperament, Puberty, and Psychosocial Behaviors.........................93 
III.  STUDY II .....................................................................................................98 
Introduction ..............................................................................................98 
Participants................................................................................................99 
Measures .................................................................................................100 
Questionnaire Measures ....................................................................100 
Temperament Questionnaire .....................................................100 
Child and Family Center Questionnaire-Child Form ...............100 
Attention Measures.................................................................................102 
General Procedures....................................................................102 
Attention Network Test .............................................................102 
Emotion and Counting Stroop Task .........................................106 
Results ....................................................................................................109 
Questionnaire Measures ...................................................................109 
EATQ-R ....................................................................................109 
Internal Consistency ...........................................................109 
Agreement between Reporters ...........................................109 
Mean Differences between Reporters ................................110 
Correlations between Scales ..............................................114 
Gender Differences .............................................................114 
Child and Family Center Questionnnaire- 
Child Form (CFCQC) ........................................................115 
Attention Measures ...................................................................116 
Attention Network Test ......................................................116 
Counting and Emotional Stroop Task ................................122 
Correlations between Attention Measures .........................126 
Adjusted Interference Scores .............................................129 
Correlations between Measures .......................................................132 
Temperament and Attention Tasks ..........................................132 
Effortful Control .......................................................................132 
Affiliativeness ...........................................................................132 
Negative Affectivity .................................................................133 
Surgency ...................................................................................133 
Adjusted Interference Scores and Father-Report 
Temperament ......................................................................133 
 
 
 ix
Summary of Correlations .........................................................141 
Temperament and Child and Family Center 
Questionnaire-Child Form (CFCQC) .................................141 
Effortful Control .......................................................................141 
Affiliativeness ...........................................................................142 
Negative Affectivity .................................................................142 
Surgency ...................................................................................147 
General Summary of Correlations ...........................................147 
Attention Tasks and Child and Family Center 
Questionnaire-Child Form .................................................147 
Multivariate Analyses .......................................................................148 
Regression Analyses to Predict Psychosocial Outcomes ................149 
Problem Behavior Composite ..................................................150 
Depressive Mood ......................................................................153 
Aggression ................................................................................154 
Prosocial Composite .................................................................155 
Parenting and Effortful Control ................................................156 
Discussion .............................................................................................157 
EATQ-R ...........................................................................................157 
Attention Tasks ...............................................................................160 
Temperament and Attention ...........................................................163 
Attention, Parenting, Risk, and Problem Behavior 
Variables ...................................................................................164 
Temperament, Parenting, Risk, and Problem Behavior 
Variables ....................................................................................165 
Temperament, Attention, Parenting Variables, and 
Psychosocial Outcomes ............................................................167 
IV.  CONCLUSIONS ........................................................................................172 
Overview ................................................................................................172 
Limitations, Future Directions, and Conclusions ............................174 
 
APPENDIX 
A.  EATQ-R SELF-REPORT ITEMS BY SCALE .......................................179 
B.  BODY CHANGES QUESTIONNAIRE - FEMALE ..............................187 
C.  STUDY 1 INSTRUCTION LETTER...................................................... [PG] 
D.  INSTRUCTION SCREENS – ATTENTION NETWORKS TASK .... [PG] 
E.  CHILD AND FAMILY CENTER QUESTIONNAIRE ........................ [PG] 
BIBLIOGRAPHY .............................................................................................. [PG] 
 x
 
LIST OF FIGURES 
Figure  Page 
1. Mean Puberty Category by Age and Gender .............................................67 
2. Mean Puberty Status Score by Age and Gender .......................................68 
3. Female Standardized Effortful Control Scale Scores 
by Pubertal Status Group .....................................................................70 
4. Female Standardized Effortful Control Subscale Scores 
by Pubertal Status Group .....................................................................71 
5. Female Standardized Surgency Scale Scores by Pubertal 
Status Group .........................................................................................72 
6. Female Standardized Surgency Subscale Scores by 
Pubertal Status Group ..........................................................................74 
7. Female Standardized Affiliation Scale Scores by 
Menarcheal Status ................................................................................75 
8. Attention Network Task Procedure .........................................................104 
9. ANT RTs by Flanker and Warning Cue Type ........................................118 
10. ANT RTs by Gender and Flanker Type ..................................................119 
11. ANT Errors by Flanker and Warning Cue Type .....................................120 
12. Counting and Emotion Stroop RTs by Stimulus Type 
and Gender .........................................................................................124 
13. Counting and Emotion Stroop Errors by Stimulus 
Type and Gender ................................................................................125 
14. Standardized Problem Behavior Scores by Levels of 
Effortful Control and Parental Monitoring ........................................156 
 
 xi
 
LIST OF TABLES 
Table                                                                                                                      Page 
1. EATQ-R Subscale Definitions and Sample Items ....................................57 
2.   Cronbach’s Alpha Scale and Subscale Reliabilities 
by Reporter, Study 1 ............................................................................61 
3.   Pearson’s Correlations between Self- and Parent-Report 
Temperament, Study 1 .........................................................................62 
4.   Temperament Scores by Gender and Reporter, Study 1 ...........................63 
5.   Pearson’s Correlations – Self-Report Temperament 
Pearson’s Correlations between Scales, Study 1 ................................65 
6.   Descriptive Statistics – Puberty Category and Pubertal 
Score by Age ........................................................................................66 
7.   Zero Order Pearson’s Correlations and Partial Correlations 
Self-Report Temperament, Females ....................................................76 
8.   Zero Order Pearson’s Correlations and Partial Correlations 
Parent-Report Temperament, Females.................................................77 
9.   Zero Order Pearson’s Correlations and Partial Correlations 
Self-Report Temperament, Males .......................................................78 
10.  Zero Order Pearson’s Correlations and Partial Correlations 
Parent-Report Temperament, Males ....................................................79 
11.  Regression Analysis to Predict Depressive Mood, Study 1 ......................83 
12.  Regression Analysis to Predict Aggression, Study 1 ................................84 
13.  Emotion and Counting Stroop Stimuli .....................................................107 
14.  Cronbach’s Alpha Scale and Subscale Reliabilities 
by Reporter, Study 2 ..........................................................................111 
15.  Temperament Score Correlations between Reporters, Sample 2 ............112 
16.  Temperament Scores by Gender and Reporter, Sample 2 ......................113 
17.  Pearson’s Correlations between Scales, Study 2 
Pearson’s Correlations – Self-Report Temperament ........................115 
18.  CFCQC Scale Scores by Gender .............................................................116 
19.  CFCQC Correlations between Scales ......................................................117 
20.  ANT RTs by Gender, Flanker Type, and Warning Cue Type ................121 
 xii
21.  ANT Attentional Network Subtractions by Gender ................................121 
22.  Counting and Emotional Stroop RTs, Conflict, and Errors 
by Gender and Stimulus Type ...........................................................122 
23.  Pearson’s Correlations – ANT Measures Genders Combined ................123 
24.  Pearson’s Correlations – ANT Measures by Gender ..............................126 
25.  Pearson’s Correlations – Counting and Emotion Stroop 
Measures Genders Combined ............................................................127 
26.  Pearson’s Correlations – Counting and Emotion Stroop 
Measures by Gender ...........................................................................127 
27.  Pearson’s Correlations between Attention Task Measures 
Genders Combined..............................................................................128 
28.  Pearson’s Correlations between Attention Task Measures 
Females ...............................................................................................128 
29.  Pearson’s Correlations between Attention Task Measures 
Male ....................................................................................................129 
30.  Total Interference Scores by Risk Group and Gender .............................131 
31.  Pearson’s Correlations between Adjusted Interference 
Scores Genders Combined .................................................................131 
32.  Pearson’s Correlations – Adjusted Interference Scores 
and Self-Report Temperament, Genders Combined .........................134 
33.  Pearson’s Correlations – Adjusted Interference Scores 
and Mother-Report Temperament, Genders Combined ...................135 
34.  Pearson’s Correlations – Adjusted Interference Scores 
and Father-Report Temperament, Genders Combined .....................136 
35.  Pearson’s Correlations – CFCQC and Temperament 
Genders Combined .............................................................................137 
36.  Pearson’s Correlations – Adjusted Interference Score, CFCQC, 
and Risk Group Gender Combined ...................................................138 
37.  Multiple Regression Analyses Dependent Variable: 
Problem Behavior Composite ............................................................139 
 
38.  Multiple Regression Analyses Dependent Variable: 
Depressive Mood ...............................................................................140 
39.  Multiple Regression Analyses Dependent Variable: 
Aggression ..........................................................................................143 
 xiii
40.  Multiple Regression Analyses Dependent Variable: 
Pro-Social Behavior Composite ........................................................144 
41.  Pearson’s Correlations – CFCQC and Temperament Females ...............145 
42.  Pearson’s Correlations – CFCQC and Temperament Males ...................146 
43.  Pearson’s Correlations – Adjusted Interference Scores, CFCQC, 
and Risk Genders Combined .............................................................151 
44.  Pearson’s Correlations – Adjusted Interference Scores, CFCQC, 
and Risk Females ...............................................................................152 
45.  Pearson’s Correlations – Adjusted Interference Scores, CFCQC, 
and Risk Males ...................................................................................152 
 
46.  [Table name] ........................................................................................... [PG] 
47.  [Table name] ........................................................................................... [PG] 
48.  [Table name] ........................................................................................... [PG] 
49.  [Table name] ........................................................................................... [PG] 
 
 
 
 
14 
 
15 
 
 
 
 
 
CHAPTER I 
 
 
INTRODUCTION 
 
 
 Adolescence has long been characterized as a time of heightened emotional 
reactivity and poor self-control.  In 1904, G. Stanley Hall characterized adolescence as a 
period of “storm and stress,” typified by questioning and contradiction of parents, mood 
disruptions, and reckless behaviors.  He posited that this period originated in "the 
chemistry" of adolescence, and represented an evolutionarily preserved “memory” of an 
earlier period of difficulty in human evolution.  The psychoanalytic viewpoint, such as 
that expressed by Anna Freud, saw adolescent storm and stress as a biologically driven 
reawakening of the Oedipal conflicts of early childhood.  This resurgence of instinctive 
drives was seen to lead to emotional volatility, depression, and antisocial behavior 
(Freud, 1968, as cited in Arnett, 1999). 
 Not all researchers have agreed with the "storm and stress" model of adolescent 
development.  For example, Margaret Mead’s (1928) anthropological studies supported 
the idea that storm and stress may be culturally driven; her work described non-Western 
cultures in which adolescents did not seem to experience emotional disruption.  Further, 
several studies in the late 60s and early 70s found that 75% of teenagers had positive 
relationships with their families (see Steinberg, 2001, for a review).  In addition, Rutter 
16 
 
and colleagues (Rutter, Graham, Chadwick, & Yule, 1976) showed that many of the 
remaining 25% had familial problems prior to the child’s entry into adolescence.  Many 
researchers subsequently dismissed "storm and stress" models of adolescence, focusing 
on adolescence as a normative period of development rather than a period of difficulty for 
adolescents and their families. 
 While researchers may have dismissed the notion that adolescence is an 
inherently challenging period of development, parents, and educators seem to have 
retained it.  Studies by Buchanan and Holmbeck (1998), among others, have found that 
the general public, including college students, teachers, and parents of adolescents, view 
early adolescence as a period of strife.  Indeed, empirical evidence also suggests general 
increases in negative affect (Buchanan, Eccles, & Becker, 1992), sensation seeking and 
risk taking behaviors (Arnett, 1999), conflict with parents (Steinberg, 2001), and 
restlessness, mood swings, and anxiety during the period of adolescence (Buchanan et al., 
1992).  These findings have led to a modified storm and stress theory, in which it is 
suggested that adolescence is “the period when ‘storm and stress’ is more likely to occur 
than at other ages” (Arnett, 1999, p. 317).  However, important individual differences 
exist in the extent to which a given adolescent will exhibit storm and stress (Arnett, 
1999). 
 The goal of the current set of studies is to examine the contribution of individual 
differences to adolescent aggression, depressive mood, and pro- and anti-social 
behaviors.  The variables examined include temperament and pubertal maturation during 
early adolescence, and temperament and executive attention, in conjunction with 
parenting variables, in later adolescence.  This approach allows examination of multiple 
17 
 
sources of influence concurrently, and may contribute to our understanding of factors that 
predict those individuals most likely to experience mood disturbances and behavioral 
difficulties during the adolescent period. 
 
Temperament, Attention, and Puberty as Individual Differences 
 
 
One important source of individual differences during adolescence is 
temperamental emotional reactivity and self-regulation.  We can think of emotional 
reactivity as encompassing the “onset, duration, and intensity of expression of affective 
reactions (e.g., fear, anger, positive affect), as well as variability in arousability and 
distress to overstimulation, activity, and attention” (Rothbart & Bates, 1998, p.108).  
Emotional reactivity is thought to arise from evolutionarily conserved affective-
motivational systems in the brain.  Individual differences in emotional reactivity can be 
seen, then, as resulting from individual differences in the neural structures and 
neurochemistry subserving such systems (Rothbart & Bates, 1998). 
 When we speak of self-regulation, we refer to the capacity for effortful control of 
action and emotion.  Effortful control involves the modulation of reactive behaviors, and 
the ability to flexibly orient attention from one focus to another.  Indeed, we view 
effortful control as reflecting functioning of a neurally based executive attention system 
that allows an individual to choose between competing response tendencies and, when 
necessary, to suppress an inappropriate response.  When older children are low in 
effortful control, their activities may continue to be driven by more reactive aspects of 
temperament, including response to reward and punishment cues in the environment 
(Rothbart & Bates, 1998).  If adolescence is inherently a time of greater than usual 
18 
 
reactivity, as suggested by Hall and others, then we may see increases in reactivity as 
being particularly difficult to manage for adolescents low in effortful control.  Indeed, 
poor self-regulation has been implicated as a risk factor in a number of studies of 
adolescent emotional dysregulation and problem behaviors (e.g., Caspi, Henry, McGee, 
Moffitt, & Silva, 1995; Dahl, 2001; White, Moffitt, Caspi, Jeglum-Bartusch, & 
Stouthamer-Loeber, 1994; Wills, Cleary et al., 2001) 
 A number of approaches have been used to assess temperament in toddlers and 
children, including parent- and teacher-report, naturalistic observations, and laboratory 
procedures designed to elicit particular behaviors (see Rothbart & Bates, 1998, for a 
discussion).  Adolescent temperament is often assessed via parent-, teacher-, and/or self-
report.  Relatively fewer observational and laboratory studies of temperament have been 
conducted with adolescents, although a number of studies have used neuropsychological 
testing batteries to assess purported functioning of frontal areas of the brain (e.g., 
Giancola, 2000; Giancola, Mezzich, & Tarter, 1998; Giancola & Tarter, 1998; Moffitt, 
1993).  In general, poor performance on these tasks has been related to anti-social 
behaviors.  We view functioning of frontal areas as vital to self-regulation; thus, these 
findings are in keeping with our theoretical model.  In addition, Krueger, Caspi, Moffitt, 
White, and Stouthamer-Loeber (1996) used a delay of gratification laboratory paradigm 
to assess self-control in 12- and 13-year-old males.  Poor delay of gratification was 
related to externalizing problems; again, this finding is consistent with our model of self-
regulation. 
Another potentially important source of individuality during adolescence is 
pubertal maturation.  While puberty is not traditionally considered an individual 
19 
 
difference variable, adolescents vary widely in the onset of, and progression through, 
puberty.  In any 7th-grade classroom, one may see individuals who are not yet showing 
any signs of pubertal development, while others appear to have completed physical 
maturation.  At any given point in time over a span of several years, pubertal maturation 
differs from adolescent to adolescent.  Pubertal maturation involves not only outwardly 
visible morphologic changes, but changes in internal neuroendocrine and CNS activity 
(Plant, 2001; Terasawa & Fernandez, 2001), making it a potentially powerful source of 
individuality during adolescence. 
The current set of studies focuses on the role of temperament, executive attention, 
and puberty in psychosocial outcomes in adolescence.  Relationships between 
temperament and puberty are also examined in early adolescence, and relationships 
between temperament and executive attention in later adolescence.  In addition, 
individual differences are examined within the context of environmental variables.  In 
this introduction, a review of the current literature on methods for assessing puberty, 
individual differences in emotional reactivity, self-regulation, attention, and pubertal 
processes is presented within the framework of adolescent psychosocial development.  
This is followed by a brief summary of studies done in our laboratory and in others that 
have contributed to the methodology and theoretical framework employed in the current 
studies.  Finally, specific hypotheses and expectations are reviewed. 
 
Assessing Puberty 
 
 
A number of methodological approaches have been developed to study the role of 
puberty in adolescent psychosocial development.  The most direct method involves the 
20 
 
measurement of pubertal hormones (androgens and estrogens), and is based on the 
assumption that levels of these hormones are directly linked to behavior.  Overall levels 
of pubertal hormones are examined via blood, urine, or saliva samples (Brooks-Gunn, 
Graber, & Paikoff, 1994).  Researchers then look for linear relationships between 
hormone levels and the behaviors of interest, typically aggression, or depressive mood. 
Several methodological difficulties are associated with this line of inquiry.  
During puberty, the hypothalamus releases gonadotropin-releasing hormone, which 
stimulates the pituitary gland to release luteinizing hormone (LH) and follicle stimulating 
hormone (FSH).  LH and FSH then stimulate the release of gonadal hormones.  The 
hypothalamic release of gonadotropic releasing hormone occurs in pulses.  These pulses 
happen primarily during sleep in the early stages of puberty, then eventually occur 
throughout the day (Plant, 2001; Terasawa & Fernandez, 2001). This pulsing action 
makes accurate measurement of hormone levels in early puberty difficult.  Levels of 
gonadotropic releasing hormone, LH and FSH, as well as androgens and estrogens, rise 
and fall during the day.  Individual differences in the number and timing of pulses 
occurring during a 24-hour period vary widely.  It is not known how quickly a pulse of 
gonadotropic-releasing hormone results in an increase in gonadal hormones (Angold, 
Costello, Erkanli, & Worthman, 1999). 
Other methods to measure pubertal development are far less complex than direct 
measurement of hormone levels.  These methods usually involve assessment of 
secondary sexual characteristics as a marker of pubertal development.  In 1969, Marshall 
and Tanner (as cited in Graber, Petersen, & Brooks-Gunn, 1996) assembled photographs 
of adolescents at various levels of pubertal development.  These photographs showed the 
21 
 
progress of breast and pubic hair growth in females, and genital and pubic hair growth in 
males.  While puberty does not involve clearly demarcated “stages,” Marshall and Tanner 
grouped photographs to represent five general levels of development: pre-pubertal, early-
pubertal, mid-pubertal, late-pubertal, and post-pubertal.  Physicians could then compare 
patients to the Tanner photographs to determine the level of pubertal development.   
These pictures have been the standard for many years in assessing pubertal status, 
and they have also been converted to drawings for use with self-assessment in children, 
as have questionnaires that ask children about their physical development (e.g., the 
Pubertal Development Scale, (Petersen, Crockett, & Richards, 1988).  These methods 
yield a puberty status score that can then be converted to a stage score matching the five 
levels of pubertal development outlined by Marshall and Tanner (as cited in Graber, 
Petersen, & Brooks-Gunn, 1996).  These pubertal status scores and stage rankings are 
often used in studies of pubertal maturation and behavior. 
Pubertal stage measures may be used in one of two ways.  The first assumes that 
the outward effects of pubertal hormones (e.g., the development of secondary sexual 
characteristics) are linked to the internal effects of pubertal hormones (e.g., 
neuroendocrine and CNS changes).  Thus, the development of secondary sexual 
characteristics serves as a proxy measure of underlying hormonal activity.  Correlations 
between Tanner stages and gonadal hormone levels range from .60 - .80 (Graber et al., 
1996), indicating that secondary sexual characteristics do provide a rough estimate of 
underlying hormonal activity.  The assessment of Tanner stages via questionnaire 
measures allows inclusion of pubertal status variables in studies where direct 
measurement of hormones or physician examination would be impossible. 
22 
 
Many studies also use assessment of Tanner stages to examine the effects of 
pubertal timing on psychosocial development.  Pubertal timing refers to the timing of 
overall pubertal development as compared to other adolescents of the same age.  An 
individual can be classified as an “early,” “on time,” or “late” maturer in comparison with 
age mates.  Pubertal timing is often determined by first establishing a pubertal status 
score for each individual within a given age group of participants.  The distribution is 
then divided to determine extreme scores.  Those with relatively low scores on pubertal 
maturation are considered “late maturers,” while those with relatively high scores are 
classified as “early maturers.”  The rest are considered as “on time.”  These 
classifications are then used to study relationships between pubertal timing and 
psychosocial development.  However, the exact methods used to derive these groups vary 
widely across studies, making comparisons of findings across studies difficult (Graber et 
al., 1996). 
Pubertal status and pubertal timing are closely interrelated, particularly in early 
puberty.  For example, measurement of pubertal status in a hypothetical group of 12-
year-old males may reveal that a third have not yet started showing signs of puberty, a 
third are showing very early signs of puberty, and a third are showing definite signs of 
pubertal development.  The first group might then be classified as “late maturers,” the 
middle group as “on time,” and the third group as “early maturers.”  By definition, then, 
pubertal status and pubertal timing would be highly correlated, because all early maturers 
would be more developed than would the late maturers.  This confound makes it difficult 
to disentangle the effects of pubertal status and pubertal timing during early adolescence. 
23 
 
Another difficulty inherent in this method is that the rate of individual progress 
through puberty varies across individuals, making classification of pubertal timing 
somewhat unstable.  A 12-year-old boy may start puberty later than his age mates, but 
complete maturation at the same time.  He would be then be reclassified from a “late 
maturer” to an “on-time maturer.”  It is not until puberty is complete that status 
differences disappear, leaving only timing differences (Graber et al., 1996), at which 
point classifications are often based on some marker of pubertal completion, such as 
menarche in females. 
Recent advances in neuroimaging suggest that puberty may affect neural systems 
thought to underlie temperament.  The hippocampus and amygdala, thought to be 
important in emotion, contain very high concentrations of gonadal hormone receptors 
(Giedd et al., 1996).  These same structures show signs of change and growth during the 
adolescent years, suggesting that these changes may be related to pubertal hormones 
(Giedd et al., 1996; Sowell & Jernigan, 1998).  In addition, areas important in executive 
functioning show substantial maturation during the adolescent period and into early 
adulthood (Klingberg, Vaidya, Gabrieli, Moseley, & Hedehus, 1999; Lewis, 1997; Paus 
et al., 1999; Sowell & Jernigan, 1998; Sowell, Thompson, Holmes, Jernigan, & Toga, 
1999; Thompson et al., 2000).  Findings such as these have prompted several current 
researchers (e.g., Dahl, 2001; Spear, 2000; Walker, 2002) to hypothesize that pubertal 
brain development activates or amplifies the activity of neurologically based emotional 
systems.  Further, Dahl (2001) suggests that frontal maturation necessary for mature self-
regulation lags behind activation of emotional systems, resulting in a period of emotional 
dysregulation.  To date, this theory has not been tested empirically. 
24 
 
 
Adolescent Emotional Reactivity and Psychosocial Development 
 
 
Negative Affectivity 
 
 
 Stereotypes of adolescent behavior often center on negative emotionality, and 
there is evidence to suggest that a general increase in negative affect does occur during 
adolescence.  Larson and his colleagues (Larson & Richards, 1996) have provided an 
illuminating look into the daily emotional lives of adolescents via a set of studies 
employing the Experience Sampling Method (ESM).  Participants carry an electronic 
pager that “beeps” them at random times throughout the day.  When a participant is 
“beeped,” he or she records his or her current emotions, thoughts, and behaviors.  Results 
indicate that adolescents experience more extreme daily highs and lows of emotion than 
do their parents, as evidenced by greater variance in scores, and greater overall negative 
affectivity than do younger children.  In fact, the amount of time experienced as “very 
happy” was 50% less in ninth-grade adolescents than in fifth graders. 
 A review of several dozen articles spanning a number of decades (Buchanan et 
al., 1992 ) reveal that a general increase in negative affect during the early adolescent 
years is a fairly consistent finding, particularly with regard to depressive or dysphoric 
mood.  In addition, some (e.g., Abe & Suzuki, 1986, as cited in Buchanan et al., 1992), 
but not all studies (e.g., Susman, Inoff-Germain, et al., 1987, as cited in Buchanan et al., 
1992) have found a peak in anxiety during the early adolescent years.  Surprisingly, even 
though adolescents are thought to be particularly irritable, few studies have attempted to 
25 
 
assess irritability in adolescents, and those that have were inconclusive (e.g., Achenbach 
& Edelbrock, 1981, as cited in Buchanan et al., 1992). 
 A recent analysis of self-reported personality has also been made from individuals 
ages 14 and older from German, British, Spanish, Czech, and Turkish samples (McCrae 
et al., 2000).  Within each cultural group, self-reported Neuroticism (a personality 
construct encompassing negative affect) declined significantly across age, with 
adolescent groups reporting significantly higher than average scores on the Neuroticism 
measure.  It is important to note, however, that no scores were reported for children; it is 
thus impossible to determine whether scores rose in adolescence or actually declined 
from even higher levels in middle childhood. 
While adolescents in general may experience increased negative affect, of equal 
interest are individual differences in temperamental negative affectivity.  In the 1950s, 
Thomas and Chess (1977) identified nine categories of infant individual differences 
based on detailed parent-report of infant behavior from parents of 22 infants.  Parents 
were asked to describe their infants’ reactions to everyday situations.  Parents’ responses 
were then sorted into the following categories: rhythmicity of biological functions, 
activity level, approach or withdrawal from novel stimuli, intensity of mood, emotional 
valence of mood, adaptability, sensory threshold, distractibility, and persistence/attention 
span.  These nine categories were then used to identify three temperamental patterns:  
easy, difficult, and slow-to-warm-up. 
 The construct of “difficult” temperament, including negative mood, withdrawal, 
low adaptability, high intensity, and low rhythmicity, also spawned a number of studies 
of early temperament.  While the dimensions included in the Thomas and Chess construct 
26 
 
of difficulty did not prove to be psychometrically sound (Putnam, Sanson, & Rothbart, in 
press), they continue to be used in a number of studies pertinent to the current review.  
 More recent studies have further refined our understanding of the emotions and 
behaviors that constitute negative affect.  Factor analytic studies of childhood 
temperament utilizing the Children’s Behavior Questionnaire have consistently revealed 
a Negative Affectivity factor consisting of scales measuring Discomfort, Fear, 
Anger/Frustration, Sadness, and low Soothability (Ahadi, Rothbart, & Ye, 1993; 
Rothbart, Ahadi, Hershey, & Fisher, 2000).  Individual differences in the affects 
contained within this broad construct show some degree of stability after the first few 
months of life.  However, while these affects tend to cluster together statistically, it has 
been argued that different negative emotions are differentially related to externalizing 
behaviors, such as aggression and anger, and internalizing problem behaviors, such as 
depression and anxiety (Eisenberg et al., 2001; Rothbart & Bates, 1998).  For example, 
Rothbart and her colleagues (Rothbart, Ahadi, & Hershey, 1994) found that laboratory 
measures of infant fear predicted childhood fear, shyness, and sadness measured via 
parent-report at 6- to 7-years of age.  However, infant frustration predicted childhood 
frustration as well as childhood approach tendencies.  Eisenberg et al., (2001) found 
concurrent relations between anger and externalizing problems and between sadness and 
internalizing problems. 
 Negative affectivity may influence psychosocial outcomes in a number of 
different ways.  Rothbart and Bates (1998) suggest that “early individual differences 
likely become transformed via developmental processes into the more complex forms of 
adjustment of later years, and these processes may shape adjustment outcomes” (p.153).  
27 
 
Lee and Bates (1985), for example, found that 2-year-old distress-prone toddlers had a 
tendency to approach trouble more often and to resist their mothers’ attempts at control.  
The mothers were more likely to use aversive discipline, but also to give in to their 
children.  Lee and Bates speculated that these patterns of mother-child interaction seen in 
toddlerhood may precede the coercive parent-child interactions related to later aggressive 
behavior (Patterson, 1980).  Bates, Pettit, and Dodge (1995) found that mother-report of 
infant difficult temperament predicted harsh parenting at age 4, which in turn predicted 
parent perception of externalizing behaviors in early adolescence.  Caspi and his 
colleagues (1995) found that early irritability and resistance to control, as assessed by 
observer ratings, predicted later externalizing problems, while early shy and inhibited 
behaviors predicted later anxiety, harm avoidance, and low aggression. 
 Concurrent relationships between negative affectivity and psychosocial 
development have also been explored.  A 1986 study of 141 adolescents and 240 college 
undergraduates (Windle et al., 1986) found significant associations between an index of 
“difficult temperament” (inflexibility, high distractibility, high intensity emotional 
responses) and both depression and perceived competence.  A one-year longitudinal 
study involving 975 adolescents found “difficult” temperament to be associated with both 
delinquency and depression, as well as higher levels of stressful life events and lower 
levels of perceived family support (Tubman & Windle, 1995).  However, findings 
regarding internalizing disorders are less clear than those examining externalizing 
disorders; overlap between measures assessing temperamental attributes such as sadness, 
and psychosocial problems such as depression, have raised the concern that the two types 
28 
 
of ratings are actually measuring one underlying pattern of behavior (Rothbart & Bates, 
1998).  This is not a problem that is easily solved, but one that deserves careful attention. 
 
Puberty and Negative Affect 
 
 
The relationship between puberty and change in negative affect has primarily 
focused on depression and aggression, rather than on changes in less pathological 
expression of negative affect.  Buchanan et al. (1992) suggest that depression and 
aggression constitute a pattern of behaviors and, as such, are easier to measure than affect 
per se.  However, they also suggest that many of the studies of behavioral change 
attribute these changes to the same neurological systems responsible for the underlying 
affect, with affect mediating the relationship between biological processes and behavior.  
As such, we feel it is appropriate to include studies examining the role of puberty in 
aggressive mood and depression under the umbrella heading of temperamental negative 
affect.  As there is currently no definitive “best way” to study the role of puberty in 
psycho-social outcomes, the studies discussed include those utilizing direct measures of 
hormone levels, as well as both pubertal status and pubertal timing as measured by 
secondary sexual characteristics. 
The speculation that depression may be related to puberty stems primarily from 
the significant gender difference in depression that arises during adolescence.  Prior to 
adolescence, males have a slightly higher rate of depression than do females.  However, 
at some point during the pubertal transition, the rate of depression in females increases 
and a 2:1, female:male ratio emerges that remains fairly consistent until menopause 
(Bebbington et al., 1998).  The exact timing of the emergence of this gender difference 
29 
 
varies from study to study (e.g., Angold, Costello, & Worthman, 1998; Bebbington et al., 
1998; Hankin et al., 1998), but it appears that this difference emerges sometime after age 
14.  Indeed, there is some evidence of increase in negative affect in 10- to 14-year-old 
girls during rapid estrogen rise (Brooks-Gunn & Warren, 1989), and a later follow up 
study showed a significant linear effect of estrogen level at Time 1 on self-reported 
symptoms of depression at time 2 (Paikoff et al., 1991), although Time 1 hormone levels 
only explained 6% of the variance in depression scores at Time 2.  The NIMH study of 
puberty and psychopathology (Susman, 1997) found lower levels of testosterone and 
higher levels of adrenal androgens to be associated with negative emotion in boys, while 
FSH correlated with negative emotion in girls. 
An additional analysis of the role of gonadal hormones in depression is provided 
by Angold and his colleagues (Angold et al., 1999).  They assessed 9-, 11-, and 13-year-
olds (N = 4500) for behavioral difficulties.  Those scoring in the top 25% on internalizing 
behaviors were then recruited for a longitudinal study, along with a non at-risk group.  
Hormonal assays were conducted, using blood drawn at two intervals, 20 minutes apart, 
then averaged.  Results indicated that females began to develop depression only above a 
particular threshold of estrogen and testosterone levels.  Tanner stage and age at 
maturation did not contribute significantly to depression scores once hormone levels were 
entered into the model.  The authors concluded that hormone levels, rather than the 
appearance of secondary sexual characteristics, were the important feature in female 
depression.  However, the authors acknowledge that all women surpass the levels of 
estrogen and testosterone implicated in the study and most do not develop depression, 
implying that a certain threshold of hormone levels is necessary, but not sufficient, for the 
30 
 
onset of depression.  Further, the authors acknowledge that their method of assaying 
hormone levels was problematic, as was their failure to standardize hormone collection 
with regard to menstrual cycle phase. 
Other evidence linking biological pubertal processes to the onset of gender 
differences in depression can be seen in a recent adolescent twin study showing that 
adolescent females’ genetic heritability for depression emerges only after pubertal onset 
(Silberg et al., as cited in Cyranowski, Frank, Young, & Shear, 2000).  Further, 
menstruation, pregnancy, and childbirth can all serve as triggers for a depressive episode 
in women vulnerable to depression, indicating a role for hormones. 
Studies examining pubertal status and internalizing problems have been largely 
inconclusive.  Several studies have failed to find a relationship between pubertal status 
and depression or depressive affect in girls (see Buchanan et al., 1992, for a review), 
while Susman et al. (1983) actually found more pubertally advanced girls to be happier 
than less pubertally advanced girls.  A 1939 study (Stone & Barker, as cited in Buchanan 
et al., 1992) found that premenarcheal girls expressed more anxiety than postmenarcheal 
girls, while Susman et al. (1983) found higher anxiety in 10- to 15-year-old boys who 
were more pubertally advanced.  While not directly related to negative affect, Susman et 
al. found that both males and females reported less energy with increasing pubertal status, 
while Buchanan et al. (1992) found pubertal status to be related to variability in energy.  
Lack of energy is often associated with depression. 
Along with pubertal hormones and pubertal status, pubertal timing has been 
implicated in models of female depression (Ge, Conger, & Elder, 1996 REF?, 2001; 
Graber, Lewinsohn, Seeley, & Brooks-Gunn, 1997).  For example, Ge, Conger, and Elder 
31 
 
(2001) followed 236 adolescent girls over a 6-year period starting in 7th grade.  During 
10th grade, girls were classified into early-, on-time, and late-maturation groups on the 
basis of self-reported age of menarche.  These groups were determined based upon a split 
in the sample of 30%, 40%, and 30%.  Early maturing girls showed the highest levels of 
depression starting in 8th grade and continuing across the span of the study.  Pubertal 
status also contributed significantly to variance in depression scores, as did negative life 
events and early depressive symptoms.  
The role of pubertal hormones in aggression has also been investigated.  In adults, 
testosterone levels tend to show concurrent relationships with aggression levels (see 
Tremblay, Schaal, Boulerice, Arseneault, & Soussignan, 1998, for a review).  This 
relationship has been observed in some studies of adolescents (Scerbo & Kolko, 1994; 
Udry & Talbert, 1988), but not in others (Mattsson et al., 1980; Susman, 1997).  In a 
three-year longitudinal study, Drigotas and Udry (1993) found that, while testosterone 
level at age 12-13 was related to both concurrent and subsequent self-reported 
delinquency, subsequent testosterone levels were not.  The authors concluded that early 
testosterone levels were related to problem behaviors through early development rather 
than hormonal effects per se.  An additional longitudinal study showed 13-year-old 
aggressive boys to have lower levels of testosterone than non-aggressive age mates 
(Tremblay et al., 1998).  However, the difference was reduced by age 14 and reversed by 
age 16, such that aggressive boys had higher levels of testosterone.  In an additional 
study, Tremblay et al. found testosterone levels to be related to social dominance, while 
body mass was related to physical aggression.  It is important to note, however, that 
32 
 
testosterone levels can be affected by social experience just prior to hormone sampling 
(Susman, 1997; Tremblay et al., 1998). 
Buchanan et al. (1992) suggest that there are too little data available to draw any 
conclusions about possible relationships between pubertal status and externalizing 
behaviors.  What evidence does exist is fairly inconclusive.  For example, Petersen and 
Crockett (as cited in Buchanan et al., 1992) found that impulse control was lowest in girls 
who were 6-12 months prior to menarche, and best in postmenarcheal girls.  Paikoff et al. 
(1991) found an increase in delinquent behavior after early puberty, while Nottelmann et 
al. (1987) found a decrease.  Finkelstein, von Eye, and Preece (1994) found a decrease in 
aggressive behavior across pubertal maturation and, while males initially reported more 
aggressive behaviors than girls, by late puberty these gender differences had disappeared.  
Brooks-Gunn, Warren, Rosso, and Gargiulo (1987) reported an interaction between 
pubertal status and life events: in the absence of negative life events, postmenarcheal girls 
were more likely to have problems, but when negative life events did occur, 
premenarcheal girls were more affected.  In a separate but related line of study, Laitinen-
Krispijn and colleagues (Laitinen-Krispijn, Van der Ende, Hazebroek-Kampschreur, & 
Verhulst, 1999) found slower progress through puberty to be associated with problem 
behaviors in both males and females participating in a study of 1300 Dutch children.  
Early studies of pubertal timing in males suggested that early maturation was 
beneficial for males (see Buchanan et al., 1992, for a review).  Early maturing males were 
generally found to feel more attractive and have better self-esteem than did later maturing 
males.  However, this effect tends to disappear over time (Graber et al., 1996).  Further, 
early maturing boys are more likely to engage in high risk behaviors (Andersson & 
33 
 
Magnusson, 1990, as cited in Ge, Conger, & Elder, 2001), and to manifest both 
internalizing distress and externalizing behaviors (Ge et al., 2001).  Further, late 
developing males may be more “flexible,” sensitive, and perceptive (Stattin & 
Magnusson, 1990).  Ge and his associates (2001) suggest that “the past undifferentiated 
view of the favorable influence of early maturation on males needs to be modified” (p. 
49) because early maturation in males has associated emotional and behavioral costs. 
One goal of the current studies is to explore the relationship between puberty and 
general negative affectivity, as well as between more regulatory aspects of temperament, 
puberty, and depressive mood and aggression.  It is hypothesized that pubertal maturation 
will show a linear relationship with increased negative affectivity, including 
temperamental frustration and the psychosocial dimensions of depressive mood and 
aggression, particularly in conjunction with poor self-regulation.  
 
Approach 
 
 
 Another basic dimension of emotional reactivity involves approach or appetitive 
behaviors.  Rothbart and her colleagues have titled this dimension of temperament 
Surgency.  Factor analytic studies of Surgency in toddlers and children reveal a pattern of 
behavior consisting of high activity and impulsivity, low shyness, and enjoyment of high-
intensity, sensation-seeking activities.  In adults, Surgency is related to the personality 
construct of Extraversion (Rothbart, Ahadi, & Evans, 2000) and is related to individual 
differences in sensation seeking in adolescents (Capaldi & Rothbart, 1992). 
 Sensation-seeking refers to the extent of an individual’s desire for novelty and 
intensity of sensory stimulation (Arnett, 1996; Zuckerman, 1988) and is associated with 
34 
 
preference for risky activities, including dangerous driving, variety of sexual experiences, 
drug and alcohol use, and minor criminal activity (Arnett, 1996).  Adolescents are often 
thought to be more extreme in sensation seeking than are children or adults, and this is 
supported by empirical evidence (Arnett, 1996; Zuckerman, 1979).  A 1996 NIDA report 
indicated that two-thirds of eighth graders have already tried alcohol and a quarter say 
that they are current drinkers.  Twenty-eight percent of eighth graders say that they have 
been drunk at least once, and heavy drinking is reported in more than 30% of high school 
seniors.  Adolescents and young adults have the highest rate of self-reported drug use 
(Kandel & Ravies, 1989) and the highest rates of sexually transmitted diseases (Irwin, 
1993).  Crime rates peak at age 18, then drop steeply (Gottfredson & Hirschi, 1990).  So 
prevalent are reckless behaviors during adolescence that Moffitt (1993) has concluded 
that “delinquency appears to be a normal part of teen life” (p. 675).  Additionally, 
adolescents and late teens report themselves as higher in the personality construct of 
Extraversion than do adults (McCrae et al., 2000). 
However, it is reasonable to suggest that individual differences in approach 
behaviors during adolescence are as likely to affect psychosocial development as are 
possible general increases.  Individual differences in sensation seeking are related to 
earlier onset and higher levels of substance use among teens (Kopstein, Crum, Celentano, 
& Martin, 2001), as well as reckless driving behaviors, risky sexual behaviors, and crime 
(Arnett, 1996).  A recent Spanish study of both institutionalized and non-institutionalized 
adolescents (Romero, Luengo, & Sobrol, 2001) found moderate relationships between 
personality and temperament measures of psychoticism, impulsivity and thrill-, 
35 
 
adventure-, and experience-seeking and overall rates of delinquency, even after 
controlling for the effects of institutionalization. 
Additional components of temperament may serve to moderate levels of 
adolescent sensation seeking.  Cloninger (1987) contends that personality is composed of 
three orthogonal components: novelty seeking, harm avoidance, and reward dependence.  
Novelty seeking is manifested in exploratory activity and intense reactions to reward, 
while harm avoidance is the tendency to avoid aversive stimuli.  Harm avoidance may be 
an important component of study in adolescent risk taking behaviors.  Wills, Windle, and 
Cleary (1998) have indeed found harm avoidance to be a protective factor in adolescents 
at high risk for substance abuse. 
Individual differences in harm avoidance may relate to individual differences in 
fear.  Fear has been posited by Rothbart and others (Gray, 1987; Rothbart & Bates, 1998) 
to work as a regulatory system that suppresses approach responses during potentially 
harmful situations.  An individual who is high in approach tendencies and low in fear 
may behave impulsively and without regard to consequences, while someone with strong 
approach tendencies who is also fearful may be better able to inhibit impulsive behaviors 
(Derryberry & Rothbart, 1997).  Indeed, infants’ fearfulness as measured in the 
laboratory negatively predicts mother-reported impulsivity, activity, and aggression at 
age 7 (Rothbart, Ahadi et al., 1994).  Further, toddler fearfulness is associated with 
greater internalized conscience in middle childhood (Kochanska, 1991, as cited in 
Rothbart, Posner, & Hershey, 1995). 
Implications of adolescent risk taking for psychosocial development are obvious 
and potentially profound.  Mortality rates increase sizably during adolescence, with 85% 
36 
 
of adolescent deaths resulting from homicides, suicides, and accidents (Irwin, 1993).  
Individuals who start drinking before age 15 have lifetime rates of alcohol dependence of 
40% (Dahl, 2001).  Contact with the legal system, contraction of a sexually transmitted 
disease, or an unwanted pregnancy can have lasting effects on an adolescent’s life.  
Moffitt, Caspi, Dickson, Silva, & Stanton, (1996) argue that risky and deviant behaviors 
during adolescence may constitute lifestyle “snares” that perpetuate deviant behavior in 
adulthood. 
While adolescent risk taking sometimes does escalate into a lifelong pattern of 
deviant behavior, the majority of adolescent risk taking is transient in nature.  Spear 
(2000) posits that adolescent risk taking and sensation seeking may be based in evolution.  
She cites evidence from animal studies indicating that “adolescents” in a variety of 
species (e.g., animals in the early stages of sexual maturation) show changes in social and 
risk taking behaviors.  She argues that these behaviors evolved as a result of the transition 
from dependence to independence that occurs during the adolescent period.  Risk taking, 
as evidenced by exploration of novel areas, may “help facilitate the dispersal of 
adolescents away from the natal family unit . . . and may have been evolutionarily 
adaptive as a means to avoid inbreeding” (p. 418).  In both rats and non-human primates, 
the period analogous to human adolescence involves exploration of new areas and 
interactions with others away from family.  In primates, males or females or both tend to 
emigrate from their natal group during adolescence, despite such emigration's association 
with considerable risk (see Spear, 2000, for discussion).  While Spear speculates that this 
adolescent increase in exploration across species is the result of maturational changes in 
37 
 
the brain, very little is currently known about physiological mechanisms that may be 
responsible for the triggering of such behaviors.  
It is hypothesized that the current study will reveal increases in temperamental 
Surgency concurrent with pubertal maturation in early adolescence.  Further, it is 
hypothesized that individual differences in Surgency will be related to problem 
behaviors, such as deviant peer affiliation and risk-taking behaviors, in later adolescence.  
These hypotheses are consistent both with existing data and with theoretical perspectives 
such as those put forth by Spear (2000).  
 
Affiliativeness 
 
 
 An additional area of interest in adolescents is that of affiliative behaviors.  
Affiliativeness encompasses the range of behaviors that encourage emotional 
connectedness in humans, such as emotional communication, intimacy, and 
responsiveness within interpersonal relationships.  Affiliativeness also includes such 
behaviors as pair-bonding and nurturance of the young in both humans and animals 
(Cyranowski et al., 2000; Frank & Young, 2000).  Gender differences in affiliative 
behaviors are evident during adolescence, with females showing a preference for 
emotional communication and intimacy in interpersonal relationships, in contrast to more 
male-stereotypical preference for independent activity and mastery.  Meta-analysis of 50 
years of personality research shows this difference to exist across culture and time 
(Feingold, 1994).  While there is evidence to suggest that gender differences exist to a 
degree prior to adolescence, Larson and Richards and their associates (Larson & 
Richards, 1989; Richards, Crowe, Larson, & Swarr, 1998) have found evidence of an 
38 
 
intensification of this difference during adolescence.  While boys in their Experience 
Sampling Method study spent an increasing amount of time alone during adolescence, 
girls spent less time alone and more time with peers than in childhood.  In addition, they 
spent more time talking than did males, and their conversations increasingly focused on 
interpersonal matters.  Girls also spent more time with the opposite sex and more time 
thinking about both opposite- and same-sex peers than did males. 
It has been suggested that females are socialized to focus on relationship 
intimacy, while males are socialized to focus more on issues of personal autonomy and 
agency (Cyranowski et al., 2000).  Intensification of gender roles may become more 
salient during the adolescent period as increased dating and opposite-sex interaction 
change adolescents’ social experiences.  It has also been suggested that biologically 
based “nurturance” systems in the brain are activated by female gonadal hormones, 
resulting in increased affiliative drives in adolescent females (Cyranowski et al., 2000; 
Frank & Young, 2000).  Animal studies have shown that the neuropeptide oxytocin plays 
a significant role in attachment and bonding behaviors (see Insel, Winslow, & Wang, 
1995, for a review).  Estrogen and progesterone receptors regulate oxytocin, and estrogen 
regulates oxytocin receptor messenger RNA.  In rats, there is a 5- to 10-fold increase in 
oxytocin mRNA at puberty.  Behavioral data from rats and voles shows that, while sex 
drive then increases in both males and females, attachment behavior only increases in 
females (see Frank & Young, 2000, for a discussion).  However, no studies to date have 
attempted to measure oxytocin in adolescent female humans; thus it is purely speculative 
at this point whether oxytocin levels and adolescent affiliative behaviors are related.  An 
increased interest in oxytocin and its role in attachment (Uvnas-Moberg, 1999), stress 
39 
 
regulation (Taylor et al., 2000), and psychosocial behaviors in adult women (Turner, 
Altemus, Enos, Cooper, & McGuinness, 1999) may spark an interest in research in 
oxytocin-behavioral relationships in adolescents. 
 Whether based in socialization or biology, individual differences in the 
intensification of affiliative behaviors in adolescence can have important ramifications 
for psychosocial development.  Adolescent relationships, both with peers and with 
romantic partners, “can have all the positive and destructive features of adult close 
relationships, and more” (Larson & Richards, 1994).  Poor peer relationships are related 
to depression and depressive symptoms in both males and females (see Petersen et al., 
1993, for a review), as is preoccupation with attachment issues (Allen, Moore, 
Kuperminc, & Bell, 1998).  Real or symbolic breaches in affiliative bonds may constitute 
a particular risk factor for depression in adolescent females.  In a recent study examining 
stressful life events in adolescent populations (Ge et al., 2001), girls with major 
depression were significantly more likely to have experienced a severely stressful life 
event during the study period than were depressed males, or control subjects of either 
gender.  Further, 68% of the depressed girls reported a stressful “interpersonal event” in 
the 6 months prior to the onset of depression, compared to 14% of the depressed boys, 
30% of the control girls, and 40% of the control boys.  Additional studies have 
underscored the importance of romantic relationships in the onset of adolescent 
depression (Darling & Cohan, 2002; Davila, Steinberg, Kachadourian, Cobb, & Fincham, 
2002; Downey & Aduk, 2002; Harper, Welsh, & Wooddy, 2002). 
 Consistent with the findings of Larson et al., as well as with the theoretical 
position of Frank & Young (2000), we hypothesize that females will show an increase in 
40 
 
affiliative desires concurrent with increased pubertal maturation.  Further, it is 
hypothesized that individual differences in affiliation will be related to depressive mood 
in both early and late adolescence, reflecting the importance of peer and romantic 
relationships in healthy psychosocial functioning. 
 
Adolescent Self-Regulation and Psychosocial Development 
 
 
Effortful Control 
 
 
While most theories of temperament stress individual differences in emotional 
reactivity, factor analytic studies of toddlers and children have revealed a voluntary 
behavioral control system that Rothbart and her colleagues have titled Effortful Control 
(Derryberry & Rothbart, 1997; Rothbart & Bates, 1998; Rothbart, Derryberry, & Posner, 
1994).  Effortful Control encompasses a broad range of volitional skills, including 
attentional, inhibitory, and activational control, and allows inhibition of a dominant 
response in order to perform a subdominant response.  Effortful Control may be of 
particular interest in the study of temperament and social development (Kochanska, 
Murray, & Harlan, 2000; Posner & Rothbart, 1998) in that it allows a child to flexibly 
and consciously direct their attention, to choose between competing response tendencies, 
and to control more reactive aspects of emotion (Putnam, Ellis, & Rothbart, 2001). 
 Voluntary, effortful control of action and emotion develops somewhat later in 
toddlerhood and childhood than involuntary, fear-based control systems.  While existing 
studies have not examined adolescent developmental changes in effortful control per se, 
adolescents in 5 different cultures report themselves as lower in the personality 
41 
 
dimension of Conscientiousness, a personality construct that encompasses some 
characteristics of effortful control, than do adults (McCrae et al., 2000).  Again, however, 
levels of child Conscientiousness are not reported, so that it is not possible to determine 
any possible change in levels of Conscientiousness from childhood to adolescence.   
 Research on individual differences in effortful control in toddlers and children has 
linked it to the development of empathy, conscience, and prosocial behaviors, and to 
lower levels of psychopathology and maladjustment (Eisenberg, 2000; Kochanska, 1997; 
Kochanska et al., 2000).  Moreover, longitudinal studies have identified several 
components of effortful control as playing an important role in adolescent psychosocial 
development.  Low self-control (as measured by poor delay of gratification) has been 
identified as a concurrent risk factor for aggressive and delinquent behaviors, while 
successful delay of gratification is linked to adaptive behaviors (Krueger et al., 1996).  
Lack of control in toddlerhood, as defined by emotional lability, restlessness, short 
attention span, and negativism, has been implicated in externalizing behaviors in 
adolescents (Caspi et al., 1995).  Lack of control at age 3 also plays a role in 
distinguishing life-course persistent antisocial offenders from those whose delinquency is 
limited primarily to the adolescent years (Moffit et al., 1996). 
 Wills and his colleagues (Wills, Cleary et al., 2001; Wills, Sandy, & Yaeger, 
2001, in press; Wills, Sandy, Yaeger, & Shinar, 2001) have identified lack of self-
regulation as a particularly strong factor in adolescent substance use and deviant peer 
associations.  Good self-control in early adolescence is concurrently related to less 
substance use by both peers and self, while poor-self control is related to deviant peer 
affiliations (Wills, Cleary et al., 2001).  In addition, good attentional orientation and 
42 
 
positive emotionality moderate the effects of parental risk on adolescent substance use 
(Wills, Sandy, Yaeger et al., 2001).  Finally, poor self-control is related to greater 
problems associated with substance use, while good self-control provides protective 
moderation (Wills et al., in press). 
 
Parenting, Peers, and Psychosocial Outcomes 
 
 
 In 1992, Patterson, Reid, and Dishion posited that parents who use coercive 
means of control, engage in frequent conflict with their children, and fail to monitor their 
children's behavior outside the home are more likely to produce aggressive, 
uncooperative children.  These children then do poorly in school and are rejected by 
mainstream peers.  This leads to association with deviant peers and eventual participation 
in deviant activities.  A number of studies have found evidence supporting this model 
(see Dishion & Bullock, 1991, for a review), with deviant peer association identified as 
the “most proximal social influence on problem behavior” (Ary, Duncan, Biglan et al., 
1999, p. 141). 
 Indeed, deviant peer association during adolescence is highly predictive of 
deviant behaviors (Dishion, Spracklen, Andrews, & Patterson, 1996), and poor parental 
monitoring is strongly associated with deviant peer involvement in early adolescence 
(Dishion, Capaldi, Spracklen, & Li, 1995).  However, Stice and Gonzales (1998), using 
both parent and adolescent measures, found that both parental control and temperamental 
behavioral undercontrol predicted unique variance in adolescent antisocial behavior and 
substance use.  Further, behavioral undercontrol moderated the relationship between 
parenting and behavior such that parenting showed stronger relations with behavior at 
43 
 
higher levels of temperamental risk.  In the current studies, we also predict that both self-
regulation and parenting behaviors will be associated with psychosocial outcomes such as 
prosocial and deviant peer association and activities.  Further, we predict that the 
combination of parenting and individual difference variables will be better predictors of 
psychosocial outcomes than either set of predictors alone. 
  
Executive Attentional Systems 
 
 
In 1986, Norman and Shallice proposed that a supervisory or executive system is 
recruited when adults are confronted by situations producing conflict between competing 
responses.  This executive system is needed in situations requiring planning, error 
correction, novel responses, difficult conditions, or in situations in which habitual 
responses must be overcome in favor of a more appropriate response (Posner & 
DiGirolamo, 1998).  Posner and Rothbart (2000) posit that executive attention underlies 
effortful control skills.  Individuals suffering damage to the frontal lobes of the brain 
often exhibit deficits in these sorts of behaviors, leading to speculation that individual 
differences in executive behaviors may be related to individual differences in functioning 
of prefrontal brain areas (Giancola & Tarter, 1998).  A variety of neuropsychological 
measures have been designed to measure executive cognitive functioning, including tasks 
that require inhibition of inappropriate responses, planning, sequencing, and word 
fluency as well as other skills (Giancola & Tarter, 1998). 
A number of studies have found relationships between poor executive cognitive 
functioning and delinquency (see Moffit & Henry, 1991, for a review).  White et al. 
(White et al., 1994) found a small, but significant relationship between performance on a 
44 
 
battery of neuropsychological tasks and delinquency in early adolescence.  These tasks 
included a time perception task, the Stroop Color and Word Association Task, the Trail 
Making Test, the Circle Tracing task, a Delay of Gratification game, and a computerized 
Card Playing task.  They also found modest relationships between performance on the 
tasks and teacher-report impulsivity.  However, while neurocognitive functioning was 
concurrently related to antisocial behavior at two time points (age 10 and age 12), age 10 
task performance did not predict age 13 delinquency, while age 10 teacher ratings of 
impulsivity did. 
 In a series of studies, Giancola and his associates (Giancola, 2000; Giancola et al., 
1998; Giancola, Moss, Martin, Kirisci, & Tarter, 1996; Giancola & Zeichner, 1994) 
found relationships between poor executive functioning, as measured by 
neuropsychological tasks such as the Wisconsin Card Sorting Test, the Porteus Mazes, 
the Tower of Hanoi, and the Stroop test, and aggression in young men (Giancola & 
Zeichner, 1994) as well as a greater degree of alcohol-induced aggression in individuals 
with deficits in frontal functioning (Giancola & Zeichner, 1994).  Preadolescents with a 
family history of substance abuse disorder also show poor frontal functioning as 
measured by these tasks when compared to controls (Giancola & Zeichner, 1994).  In one 
of the few studies to specifically examine frontal functioning in adolescent females, poor 
performance on neuropsychological tasks predicted antisocial behaviors as hypothesized 
(Giancola et al., 1998), suggesting that poor executive functioning is associated with poor 
psychosocial outcomes across genders.  In 15- to 18-year-old males, Giancola found a 
significant relationship between poor executive functioning and negative affectivity, both 
of which were higher in youths at familial risk for substance use disorders.  While poor 
45 
 
executive functioning was associated with drug use in the sample studied, this 
relationship was only evident in individuals with a family history of substance abuse, 
suggesting that individuals with poor executive functioning in combination with family 
risk factors may be especially vulnerable to developing substance use disorders. 
Performance on measures of neuropsychological functioning such as the 
Wisconsin Card Sorting Test, California Verbal Learning Test-Children's Version, Word 
Fluency, Animal Naming and Desing Fluency, the Twenty Questions task, the Go/No Go 
task, the Tower of London, and Delayed Alternation task, has been shown to improve 
from childhood through adolescence (Levin, Culhane, Hartmann, Evankovich, & et al., 
1991).  Specifically, Levin et al. (1991) found majors gains between a 7- to 8-year-old 
group and a 9- to 12-year-old group on the Wisconsin Card Sort and false-positive errors 
on the Go/No Go Task.  However, further advances in performance in 13-15 year olds 
were confirmed in a group of 13-15 year olds for the remained of the tasks. 
In addition, Williams, Ponesse, Schachar, Logan, and Tannock (1999) found 
evidence for increased ability to inhibit a response during a Go/No Go task into 
adulthood.  It has been suggested, however, that maturation of executive functioning may 
be different in individuals with antisocial behavior disorders.  A 1999 cross-sectional 
study of normally developing and behaviorally disturbed children ages 6-16 found 
improved competence on measures of executive functioning across age, regardless of 
behavioral difficulties (Mezzacappa, Kindlon, & Earls, 1999).  However, children with 
externalizing disorders showed lower levels of executive functioning at all ages than did 
non-disordered children, indicating a stable deficit in frontal functioning. These findings 
46 
 
suggest that the relationship between executive functioning and psychosocial outcomes 
may be present, and measurable, early in life. 
In the current set of studies, we predict that effortful control variables, as 
measured both by self- and parent-report, will be related to performance on a newly 
developed measure of executive attention task in older adolescence, as well as to 
performance on a measure of executive attention designed to tap both cognitive and 
emotional attention.  Further, we predict that both questionnaire measures of self-
regulation and performance on executive attention tasks will predict psychosocial 
outcomes.  Further, we predict that parenting and self-regulation variables will contribute 
both uniquely and jointly to the prediction of psychosocial outcomes, including both 
prosocial and problem behaviors. 
 
Preliminary Studies 
 
 
Evidence for an important role of temperamental individual differences in 
adolescent psychosocial development suggests that this line of study is an important one.  
This prompted our 1998 project to revise and update the Early Adolescent Temperament 
Questionnaire (EATQ; Capaldi & Rothbart, 1992; Ellis & Rothbart, 2002).  While the 
EATQ has been shown in two separate studies to be a valid and reliable measure of many 
aspects of adolescent temperament, we felt that it did not contain sufficient scales to 
measure fully the subcomponents of effortful control.  Thus, we added scales designed to 
do so, and embedded scales measuring aggression and depressive mood within the 
instrument to examine possible relationships between temperament and social-emotional 
functioning. 
47 
 
We administered the questionnaire to 177 school students (82 males, 94 females, 
1 unidentified), aged 10-15 years.  In addition, a parent-report version was completed by 
62 parents of adolescent participants.  Our examination of the relation of temperament to 
psychosocial behavior underscored our belief that temperament, and particularly effortful 
control, plays an important role in adolescence.  Multiple regression analyses revealed 
that low Effortful Control, high Surgency, and low Affiliativeness best predicted 
aggression scores, while low Effortful Control, high Affiliativeness, high Frustration, and 
gender (i.e., being female), best predicted depressive mood scores.  In both cases, 
examination of standardized beta weights revealed Effortful Control to predict the 
greatest amount of variance. 
Recent research has sought to relate the study of effortful control to the operation 
of high-level attention networks (Posner & Rothbart, 2000).  One widely used measure of 
the ability to resolve cognitive conflict is the Stroop task.  Stroop tasks require that 
participants name the ink color of a printed word, while ignoring the content of the word.  
In compatible conditions, ink color matches the meaning of the printed word (e.g., “red,” 
printed in red ink), while incompatible conditions involve ink color that is different than 
the meaning of the word (e.g., “red,” printed in green ink).  The latter condition induces 
conflict, and requires participants to suppress automatic processing of the semantic 
content of the word to name the ink color, resulting in longer reaction times (RTs). 
The neural networks involved in such tasks have been studied fairly extensively in 
adults.  Conflict tasks as diverse as the color-word Stroop effect (Bush, Luu & Posner, 
2000), the numerical Stroop task (Bush et al., 1998), and the use of congruent and 
incongruent flankers (Botvinick, Nystrom, Fissell, Carter, & Cohen, 1999) have all 
48 
 
shown activity in the anterior cingulate cortex (ACC), often in conjunction with lateral 
frontal areas (Bush, Luu & Posner, 2000).  Imaging studies have also revealed a failure of 
ACC activation in adults with ADHD during a counting Stroop task (Bush et al., 1999). 
Emotional Stroop-type tasks, involving presentation of negatively valenced versus 
neutral words, also activate the anterior cingulate, in an area of ACC that connects to the 
amygdala, hypothalamus, hippocampus and other areas important in regulation of 
emotional responses (Bush, Luu, & Posner, 2000).  Measures of this type have been used 
extensively in studies investigating cognitive bias in psychopathology.  A 1996 review of 
these studies (Williams, Matthews, & MacLeod) concluded that, in general, patient 
populations suffer cognitive interference, resulting in poor task performance, when 
presented with emotional stimuli related to their current condition.  This effect has been 
studied in numerous disorders, such as PTSD, depression, and anxiety in adults (e.g., 
McNeil, Tucker, Marand, Lewin, & Nordgran, 1999), depression and anxiety in 9- to 18-
year-olds (Doost, Taghavi, Moradi, Yule, & Dalgleish 1997), PTSD in children and 
adolescents (Moradi, Taghavi, Doost, & Dagleish, 1999), and anxiety in adolescents 
(Richards, Richards, & McGeeney, 2000).  These findings suggest that executive 
attention may also play an important role in the development of internalizing disorders, 
perhaps due to an inability to shift attention from emotionally charged stimuli. 
Drs. Rothbart and Posner have used neural marker tasks, shown by imaging 
studies to activate particular regions of the brain, to trace the early development of 
attentional networks and to relate this development to caregiver-report of temperamental 
effortful control and affect.  Gerardi-Caulton (2000) developed a Stroop task appropriate 
for use with children as young as 24 months-of-age.  The task introduced conflict 
49 
 
between the location of an event and its identity.  Stimuli were presented on either the left 
or right side of a computer monitor.  On either side of the monitor were two large 
“buttons” containing a picture.  The children were instructed to touch the button that 
matched the picture on the screen.  In the compatible condition, the picture appeared on 
the same side of the screen as the correct button.  In the incompatible condition, the 
picture appeared on the side of the screen opposite the correct button. 
There is a strong tendency to respond to the location of the stimulus, when put in 
conflict with the identity.  When the correct key is on the opposite side, a conflict 
emerges and even adults show longer reaction times (RTs) and higher error rates on these 
conflict trials.  It was found that children of two years were almost always able to make 
the correct response on compatible trials but were almost always incorrect on 
incompatible trials.  Three-year-olds, however, made fewer errors although, like adults, 
they had longer RTs on incongruent trials.  This finding has since been replicated in a 
cross-sectional study of toddlers 24-, 30-, and 36-months of age (Rothbart, Ellis, Rueda, 
& Posner, 2002). 
Gerardi-Caulton (2000) found that performance in the spatial conflict task was 
related to effortful control, as measured in both the laboratory and via parent report 
temperament scales.  The recently completed cross-sectional study (Rothbart et al., 2002) 
also found a relation between performance on the spatial conflict task and caregiver 
reports of temperamental effortful control.  Additionally, both studies found that children 
better able to resolve the conflict introduced by spatially incompatible trials had lower 
scores on caregiver-reported negative affectivity.  
 
50 
 
The Current Studies 
 
 
As described previously, a number of recent studies have suggested that areas of 
the brain important in executive functioning, as well as areas important in emotion, 
undergo significant development during the adolescent years (Giedd et al., 1996; 
Jernigan, Trauner, Hesselink, & Tallal, 1991; Klingberg et al., 1999; Paus et al., 1999; 
Sowell & Jernigan, 1998; Sowell et al., 1999).  It has been further suggested that this 
development plays a role in adolescent behavior, self-regulation, and the onset of 
psychopathology (Dahl, 2001; Spear, 2000; Walker, 2002).  Individual differences in 
temperamental emotional reactivity and self-regulation, as well as performance on tasks 
related to neural systems thought to underlie self-regulation, have been identified in 
previous studies as important in adolescent psychosocial outcomes.  These findings, in 
conjunction with the temperament and attention studies in our laboratory, prompted 
development of the current set of studies. 
In Study 1, we sought to relate the physiological changes associated with puberty 
to increases in temperamental reactivity in a community sample of early adolescents.  
Because it was not feasible to measure directly hormone levels or brain changes, we used 
secondary sexual characteristics as a proxy measure of underlying hormonal activity.  We 
hypothesized that emotional reactivity, encompassing surgent behaviors and frustration, 
depressive mood and aggression, would be higher in more pubertally advanced early 
adolescents.  Further, we predicted that pubertal maturation combined with relatively 
poor effortful control would best predict scores on depressive mood and aggression.  
Based on our prior research (Ellis & Rothbart, in preparation), it was also predicted that 
51 
 
approach tendencies, or surgency, would be related to aggression but not to depressive 
mood, while affiliativeness would predict depressive mood, but not aggression. 
In Study 2, we utilized an ethnically diverse sample of 16- and 17-year-olds, in 
which one subgroup had been identified in middle school as being “at risk,” and another 
as “high risk” for the development of anti-social behaviors.  Participants were asked to 
report on family variables such as parental control, rules, and relationship quality, as well 
as their own involvement with prosocial and anti-social peers.  We sought, first, to relate 
parent- and self-report effortful control to performance on measures of both cognitive and 
emotional conflict.  Specifically, we predicted that individuals with relatively poor 
Effortful Control would show relatively greater interference in these tasks.  In addition, 
we predicted that overall attention interference would be positively related to negative 
affectivity and that emotional conflict would be related to higher levels of depressive 
mood.  Second, we sought to better understand the contributions of temperament and 
executive attention to aggression, depressive mood, and both pro- and anti-social peer 
affiliation and behaviors.  This allowed examination of individual differences within the 
context of risk category, parenting, and family relationship variables.  Specifically, we 
predicted that risk category and family variables would contribute to psychosocial 
outcomes, but that prediction would improve significantly with the addition of individual 
difference variables. 
A secondary focus of both studies was to replicate and extend our previous 
findings (Ellis & Rothbart, 2002) with regard to the psychometric properties of the 
EATQ-R, as well as to explore areas of agreement and disagreement between self- and 
parent-report in two samples varying in age and background.  The set of studies provided 
52 
 
an excellent opportunity to assess the EATQ-R as an instrument appropriate for use 
across a wide range of ages, ethnic groups, and risk groups. 
The following chapters detail the methodology and results of each study in turn, 
including first a discussion of descriptive findings and psychometrics of the instruments, 
followed by discussion of the primary hypotheses for the specific study.  This is followed 
by a more general discussion that includes limitations of the current studies and 
suggestions for future research.
53 
 
 
 
 
 
 
 
 
CHAPTER II 
 
 
STUDY I 
 
 
Introduction 
 
 
 The primary goals of Study 1 were to explore the relation between pubertal 
maturation and temperament during early adolescence, and to evaluate the role of 
individual differences in both temperament and pubertal status in depressive mood and 
aggression.  Specifically, it was hypothesized that emotional reactivity would increase 
with pubertal maturation as suggested in previous studies and by recent theorists.  
Further, it was hypothesized that both temperament and pubertal maturation would 
contribute significantly to explanation of scores on depressive mood and aggression.  
Secondarily, Study 1 sought to replicate and extend the findings of our previous study of 
adolescent temperament (Ellis & Rothbart, in 2002) on the psychometric properties of the 
EATQ-R and agreement between parent and adolescent reporters. 
  
Participants 
 
 
 Participants in Study 1 were 165 adolescents (mean age = 12.31; SD = 1.58) and 
his or her primary caregiver.  Caregiver respondents were assumed to be primarily 
54 
 
mothers of participants; however, parent participants were not asked to report their 
gender.  Seventy-seven females (mean age = 12.38; SD = 1.70) and 71 males (mean age = 
12.24; SD = 1.25) participated.  One female participant failed to answer approximately 
one-half of the questions asked; her responses were not included in analyses.  Participants 
were not asked about their ethnic backgrounds and the investigator did not meet the 
participants; however it is assumed that the majority of participants were European-
American, reflecting the demographic composition of Eugene and the surrounding area.  
Participants were recruited by telephoning parents/caregivers listed in a database derived 
from birth announcements published in a local newspaper.  Adolescent participants 
received a $5.00 gift certificate redeemable at a local mall as a token of appreciation for 
participation.  Parent/caregiver participants did not receive compensation for 
participating. 
 
Measures 
 
 
Temperament Questionnaire 
 
 
In 1998, we (Ellis & Rothbart, 2002) initiated a project to revise the Early 
Adolescent Temperament Questionnaire (EATQ; Capaldi & Rothbart, 1992).  A number 
of important changes were made in an effort to develop an instrument appropriate for use 
in studies of development and psychosocial outcomes.  First, we added scales designed to 
measure Effortful Control.  Second, we included scales designed to measure depressive 
mood and aggression.  However, due to conceptual overlap between items assessing 
sadness and depressive mood, and items assessing anger and aggression, scales 
55 
 
measuring sadness and anger were not included.  Third, we developed a scale intended to 
measure affiliative behaviors. 
We administered the revised instrument to 176 adolescents between 10- and 15-
years of age.  In addition, 62 parents filled out a parent-report form of the instrument.  
We then conducted analyses assessing scale reliability, dimensionality, and discriminant 
validity, and made further revisions by excluding scales that did not achieve acceptable 
coefficient alpha levels.  Coefficient alpha for the remaining scales ranged from .65 - .82.  
Convergence with parent report was modest for most scales, but poor for depressive 
mood in both genders and, in boys, for shyness and inhibitory control.  We performed 
exploratory factor analyses of the temperament scales and found four clear factors:  
Effortful Control, made up of scales measuring Attention, Inhibitory Control, and 
Activation Control; Surgency, including scales measuring High Intensity Pleasure and, 
loading negatively, Shyness and Fear; Negative Affectivity, including the Frustration 
scale as well as the two socio-emotional scales, as well as a clear Affiliativeness factor 
comprised of scales measuring Affiliation, Perceptual Sensitivity, and Pleasure 
Sensitivity. 
With the exception of the emergence of the Affiliativeness factor, our findings 
were consistent with previous research on the broad constructs of temperament in 
childhood; however, factor loadings were somewhat different, particularly with regard to 
the loading of Fear on the Surgency factor rather than with Negative Affectivity.  We 
then took 4-9 items from each scale that loaded on a factor, and combined them to make 
one broad scale containing separate subscales.  This procedure was repeated with the 
parent-report form.  This allowed us to create a relatively short instrument that still 
56 
 
contains the breadth needed to measure important aspects of temperament such as 
Effortful Control. 
This shortened instrument became the Early Adolescent Temperament 
Questionnaire - Revised (EATQ-R), which contains 65 questions in the self-report form, 
and 62 questions in the parent-report form.  The self-report form asks adolescents how 
true each statement is for them; parent-report form asks the parent/caregiver to 
characterize the truth of each statement as it refers to his or her child.  Response options 
for both forms used a 5-point Likert-type scale:  1 = Almost always untrue; 2 = Usually 
untrue; 3 = Sometimes true, sometimes untrue; 4 = Usually true; and  5 = Almost always 
true. 
Both forms include scales designed to measure temperamental Effortful Control, 
Affiliativeness, Surgency, and Negative Affectivity.  However, the parent-report form 
does not include Perceptual Sensitivity and Pleasure Sensitivity subscales within the 
Affilativeness scale.  These subscales were not included because they contain items that 
would be very difficult for an observer to judge about another person (e.g., “I tend to 
notice little changes that others do not notice.”).  Table 1 contains definitions and sample 
items for each subscale, while Appendix A contains a full listing of items by scale. 
 
Scoring 
 
 
 Appropriate items are reverse scored, then subscale scores are calculated as a 
mean score of applicable scale items.  The scales are scored so that a high score indicates 
that the assessed dimension is high for that individual.  Missing items are not included in 
57 
 
TABLE 1.  EATQ - R Subscale Definitions and Sample Items 
Scale Subscale Definition and Sample Item 
Effortful Activation   The capacity to perform an action when there is a strong 
Control Control tendency to avoid it.  “If I have a hard assignment to do, I get started right away.” 
The capacity to focus attention as well as to shift attention 
 Attention when desired.  “I pay close attention when somebody tells me 
how to do something.” 
 Inhibitory The capacity to plan, and to suppress inappropriate responses.  Control “It’s easy for me to keep a secret.” 
The desire for warmth and closeness with others, independent 
Affiliativeness Affiliation of shyness or extraversion.  “It is important to me to have 
close relationships with other people.” 
Perceptual Detection or perceptual awareness of slight, low-intensity  Sensitivity stimulation in the environment.  “I tend to notice little changes that other people do not notice.” 
Pleasure Amount of pleasure related to activities or stimuli involving  Sensitivity low intensity, rate, complexity, novelty, and incongruity.  “I like the crunching sound of autumn leaves.” 
High Intensity The pleasure derived from activities involving high intensity  Pleasure or novelty.  “I wouldn’t be afraid to try something like mountain climbing.” 
Surgency Fear (reverse Unpleasant affect related to anticipation of distress.  “I worry scored) about getting into trouble.” 
Shyness 
 (reverse Behavioral inhibition to novelty and challenge, especially 
scored) social.  “I am shy about meeting new people.” 
Negative affect related to interruption of ongoing tasks or goal 
 Frustration blocking.  “I get irritated when I have to stop doing something 
I’m enjoying.” 
Negative Depressive Unpleasant affect and lowered mood, loss of enjoyment and Effect Mood interest in activities.  “My friends seem to enjoy themselves more than I do.” 
 Hostile and aggressive actions, including person- and object-
Aggression directed physical violence, direct and indirect verbal aggression, and hostile reactivity.  “I pick on people for no 
real reason.” 
 
 
calculation of the mean.  Scale scores are calculated as a mean of appropriate subscale 
scores; for example, Effortful Control scores consist of the mean of the Attention, 
Inhibitory Control, and Activation Control subscale scores.  In the case of Surgency, 
58 
 
subscale scores for Fear and Shyness are reversed prior to calculation of the Surgency 
mean score. 
 
Pubertal Status 
 
 
 Pubertal status was assessed via parent- and self-report versions of the Body 
Changes Questionnaire, a scale adapted from Carskadon and Acebo (1993) and based on 
Petersen et al.’s (Petersen, Crockett, Richards, & Boxer, 1988) interview-based Pubertal 
Development Scale.  The Pubertal Development Scale (PDS) was developed for use in 
studies for which direct measures of pubertal status, such as physician assessment, is not 
feasible.  A three-year longitudinal study (Petersen, Crockett, Richards et al., 1988) found 
PDS scales to be internally reliable (coefficient alpha: .68 - .83, median alpha = .77).  
Correlations between PDS questionnaire reports and physician ratings for female subjects 
(Brooks-Gunn et al., 1987) were generally high (.61 - .67).  In the current version, both 
males and females were asked to report on pubertal development of body hair growth and 
skin changes; males were also asked about voice changes and facial hair growth and 
females were asked about breast growth and menstruation.  Most questions had response 
options ranging from 1 - “Hasn’t started” to 4 - “Seems complete.”  In addition, 
respondents could answer that they didn’t know.  Female menstrual status was assessed 
via a yes or no response, and if “yes,” age of menarche was also assessed.  Caregiver 
forms were modified such that questions referred to “your child.”  The original 
Carskadon and Acebo scale also contained an item assessing growth spurt; the current 
study did not include an item assessing growth spurt.  The complete questionnaires are 
included in Appendix B. 
59 
 
 
Scoring 
 
 
 All questions for males and questions 1 to 3 for females were scored as follows: a 
= 1, b = 2, c = 3, d = 4, and 5 = no score given.  Question 4 for females was scored as 
follows:  “no” = no score given, “yes” = 4.  Overall scores for both self- and parent- 
report were calculated as a mean of all items for which a score was given.  A composite 
score was calculated that represented the mean of parent- and self-report scores. 
In addition, each participant was assigned a Puberty Category Score, using a 
modification of procedures used by Carskadon and Acebo (1993).  For self- and parent-
report male forms, responses for questions 1, 3, and 4 (body hair, voice change, and facial 
hair) were added together, and a composite score calculated representing the mean of 
parent- and self-report.  This provided a multiple respondent score rather than a single 
respondent score as calculated by Carskadon and Acebo.  Categories were determined by 
composite score point values:  Prepubertal = 3, Early Pubertal = 3.5 - 5, Midpubertal = 
5.5 - 8, Late pubertal = 8.5 – 11, and Postpubertal = 12.  For both self- and parent-report 
female forms, responses for questions 1 and 3 (body hair and breast growth) were added 
together, and a composite score was calculated that represented the mean of parent-and 
self-report.  Categories were determined by composite score values in addition to 
menstrual status:  Prepubertal = 2, Early Puberty = 2.5 – 4, Mid-pubertal = 4.5 - 6 and no 
menarche, Late pubertal = 7 or less and menarche, Postpubertal = 8 and menarche. 
 
60 
 
General Procedure 
 
 
 After being contacted by phone and giving initial agreement to participate, 
caregivers were mailed packets containing all parent- and self- report questionnaires, as 
well as consent forms for parent participants and assent forms for adolescent participants.  
An instruction letter specified that parent and adolescent participants should complete the 
questionnaires independently (see Appendix C).  A plain envelope was provided in which 
adolescent participants were instructed to seal their completed questionnaires.  This 
envelope was then included with completed caregiver questionnaires and returned to the 
principal investigator in a self-addressed, stamped envelope. 
 
RESULTS 
 
 
Descriptive and Psychometric Properties of the EATQ - R 
 
 
A series of item-total analyses were performed for subscales and scales by reporter.  
Coefficient alpha (Cronbach’s) was estimated for each scale and subscale.  For self-report 
scales, alpha ranged from .67 - .80 and alpha for subscales ranged from .55 - .78 (see 
Table 2).  For parent-report, scale alphas ranged from 70 - .87 and alpha for subscales 
ranged from .56 - .82.  Average inter-item correlation was calculated for each subscale to 
further insure internal consistency even within small clusters of items covering broad 
dimensions of behavior.  Subscale mean inter-item correlations averaged .37.  
 
Agreement between Reporters 
 
 
Convergence between self- and parent-report was assessed using a series of 
61 
 
Pearson’s correlations.  Agreement between self- and parent-report scores by gender and 
total is reported in Table 3.  Convergence for males and their parent ranged from .22 - .65 
 
 
TABLE 2.  Cronbach’s Alpha Scale and Subscale 
Reliabilities by Reporter, Study 1 
 
Scale Subscale Self Report Parent Report 
Effortful Control  .80 .87 
 Activation Control .73 .81 
 Attention .62 .73 
 Inhibitory Control .57 .56 
Affiliativeness  .76 - 
 Affiliation .66 .70(a) 
 Perceptual Sensitivity .66 - 
 Pleasure Sensitivity .77 - 
Surgency  .67 .80 
 High Intensity Pleasure .61 .74 
 Fear (reverse scored) .55 .74 
 Shyness (reverse scored) .78 .90 
Negative Affect  .79 .88 
 Frustration .72 .79 
 Depressive Mood .64 .75 
 Aggression .74 .82 
aOnly the Affiliation subscale is included in parent report Affiliativeness 
 
 
and agreement was statistically significant for all scales and subscales with the exception 
of Inhibitory Control, which was significant at trend level.  Agreement between females 
and their parent ranged from .22 - .51, and was statistically significant or marginally 
significant for all scales and subscales.  With genders combined, agreement was 
statistically significant for all scales and subscales, ranging from .29 - .53. 
62 
 
 
TABLE 3.  Pearson’s Correlations between Self- and 
Parent-Report Temperament, Study 1 
 
 Male       Female    
n = 70 n = 76 Total 
Effortful Control .54** .41** .50** 
Activation Control .65** .37** .53** 
Attention .35** .42** .42** 
Inhibitory Control .22+ .22+ .29** 
Affiliation .36** .38** .42** 
Surgency .40** .48** .46** 
High Intensity Pleasure .54** .37** .48** 
Fear .37** .27* .33** 
Shyness  .33** .51** .47** 
Negative Affect .42** .38** .38** 
Frustration .36** .23* .30** 
Depressive Mood .34** .28* .33** 
Aggression .43** .40** .41** 
+ p < .10; * p < .05; ** p < .01 
 
 
Scale means by gender and reporter are presented in Table 4.  Female participants 
consistently rated themselves higher on Effortful Control subscales than did their parent 
(Activation Control; t(75) = 3.27, p < .01; Attention; t(75) = 2.53, p < .01; Inhibitory 
Control; t(75) = 1.83, p = .07, total Effortful Control; t(75) = 3.37, p < .001).  In addition, 
they rated themselves higher on the Fear subscale (t(75) = 3.08, p < .01) and lower in 
overall Surgency (t(75) = -3.47, p < .01).  Further, females reported themselves higher in 
63 
 
Frustration than did their parent (t(75) = 2.19, p < .05), and lower in Aggression (t(75) =  
-6.40, p < .001). 
 
TABLE 4.  Temperament Scores 
by Gender and Reporter, Study 1 
 
Self, Mother, Self, Mother, 
Scale Female Female Male Male 
n = 76 n = 76 n = 70 n = 70 
M (SD) M (SD) M (SD) M (SD) 
Effortful Control 3.74(.53) 3.52(.54)** 3.57(.57) 3.38(.57)** 
Activation   
Control 3.67(.75) 3.37(.71)**
  3.54(.88) 3.09(.80)** 
Attention 3.62(.56) 3.42(.69)**  3.47(.66) 3.30(.63)*  
Inhibitory 
Control 3.92(.66) 3.77(.53)*
  3.69(.63) 3.73(.58) 
Affiliation 3.99(.63) 4.03(.49) 3.36(.76) 3.74(.83)** 
Surgency 3.26(.57) 3.47(.49)**  3.36(.52) 3.48(.57)+  
High Intensity 
Pleasure 3.14(.79) 3.43(.65) 3.27(.71) 3.37(.73) 
Fear  2.92(.75) 2.65(.62)**  2.66(.73) 2.34(.69)** 
Shyness  2.43(.98) 2.37(.82) 2.53(.95) 2.56(.97) 
Negative Affect 2.52(.48) 2.62(.59) 2.68(.82) 2.58(.55) 
Frustration 3.31(.71) 3.10(.69)*  3.37(.67) 2.99(.67)** 
Depressive 
Mood 2.49(.64) 2.48(.70) 2.48(.69) 2.32(.67)+
  
Aggression 1.76(.60) 2.31(.72)**  2.19(.68)  2.42(.66)** 
 Paired Samples t test 
 Difference between parent- and self-report: + p < .10, * p < .05, ** p < .01. 
 
 
Males also rated themselves higher in Activation Control (t(70) = 5.05, p < .001), 
Attention (t(70) = 2.12, p < .05) and total Effortful Control (t(70) = 3.04, p < .01) than 
did their parent, but not Inhibitory control.  Males reported themselves as lower in 
64 
 
Affiliation (t(70) = -3.57, p < .01), higher in Fear (t(70) = 3.30, p < .01; overall Surgency; 
(t(70) = -1.77, p = .08), higher in Frustration (t(70) = 4.23, p < .001), and lower in 
Aggression (t(70) = -2.70, p < .01) than did their parent. 
 
Gender Differences 
 
 
 Gender differences were also examined.  Females reported themselves as 
significantly higher on subscales measuring Affiliation (t(144) = -5.34, p < .001), 
Pleasure Sensitivity (t(144) = -3.39, p < .001), and Fear (t(144) = -2.00, p < .05) than did 
males; and higher in Inhibitory Control at trend level (t(144) = -1.95, p = .053).  Self 
reported Aggression was significantly higher for males in this sample than for females 
(t(144 = 3.91, p < .001).  Gender differences in scale scores reached significance for 
Affiliativeness (females higher, t(144) = -3.83, p < .001), and trend level for both 
Effortful Control (females higher, t(144) = -1.76, p = .08) and Negative Affectivity 
(males higher, t(144) = 1.89, p = .06). 
 
Correlations between Temperament Scales and Socio-Emotional Scales 
 
 
 Correlations between temperament scale scores are presented in Table 5, with 
parent-report correlations presented above the diagonal, and self-report presented below.  
In both sets of reporters, Effortful Control was negatively related to Frustration, 
Aggression, and Depressive Mood.  Depressive Mood was negatively related to Surgency 
and positively related to Aggression and Frustration in both parent- and self-report.  
Aggression was positively related to Frustration and negatively related to Affiliation in 
both parent- and self-report.  Frustration and Surgency were also negatively related in 
65 
 
both sets of reporters.  In addition, parent-report Effortful Control was related to parent-
report Surgency.  This relationship was not evident in self-report; however, self-report 
Effortful Control and Affiliation were positively related at trend level.  Parent-report 
Depressive Mood was negatively related, at trend level, to Affiliation.   
 
 
TABLE 5.  Pearson’s Correlations – Self-Report Temperament 
Pearson’s Correlations between Scales, Study 1 
 
 1 2 3 4 5 6 
1-Effortful Control - .04 .24** -.40** -.35** -.46**
2-Affiliation .15+ - .27** -.27** -.16+ -.09 
3-Surgency .07 .13 - -.19* -.50** -.23**
4-Aggression -.58** -.19* -.02 - .35** .65** 
5-Depressive Mood -.48** .07 -.32** .31** - .51**
6-Frustration -.28** -.02 -.25** .31** .36** - 
Parent –report values above the diagonal, self-report values below 
+ p < .10, * p < .05, ** p < .01 
 
 
 
Body Changes Questionnaire 
 
 
 Means and standard deviations for both pubertal status scores and puberty 
categories by age and gender are presented in Table 6, and are illustrated in Figures 1 and 
2.  Eighteen participants (4 female) were classified as prepubertal, 66 (26 female) as early 
pubertal, 44 (31 female) as mid-pubertal, and 18 (15 female) as late pubertal.  No 
participants reached scoring criteria to be classified as post-pubertal.  
66 
 
Age and Gender Differences 
 
To investigate age and gender differences in Puberty Category scores, a univariate 
analysis of variance was performed, with age and gender entered as independent 
variables.  Significant main effects of both age (F(4, 136) = 23.05, p < .001) and gender 
(F(1, 136) = 20.89, p < .001) were present, indicating an increase in scores across age for 
 
 
TABLE 6.  Descriptive Statistics – Puberty Category 
and Pubertal Score by Age 
 
 10 11 12 13 14 and older Total
n 6 60 37 15 28 146
Females 3 34 16 8 15 76
Males 3 26 21 7 13 70
Mean 
Puberty 1.50(.55) 2.17(.69) 2.14(.67) 2.87(.74) 3.32(.77) 2.43(.86)
Category 
Females 1.67(.57) 2.44**(.66) 2.44*(.63) 3.25*(.71) 3.73*(.46) 2.75(.84)**
Males 1.33(.58) 1.81(.57) 1.90(.63) 2.43(.54) 2.85(.80) 2.07(.75)
Mean 
Pubertal 3.08(.86) 4.34(1.21) 4.43(1.27) 5.57(1.06) 7.27(1.47) 5.00(1.73)
Score 
Females 2.67(.77) 4.50(1.33) 4.53(1.42) 5.75(1.14) 7.36(.95) 5.13(1.74)
Males 3.50(.87) 4.15(1.02) 4.38(1.17) 5.36(1.03) 7.15(1.71) 4.86(1.71)
Independent Samples T-Test, Gender Difference:  *p<.05, **p<.01 
 
 
 
both genders, as well as higher scores at each age for females than for males.  Pubertal 
Status scores could not be compared across genders as scores were calculated using 
different criteria;1 therefore one-way ANOVA was conducted separately for each gender.  
                                                 
1 Males receive 1, 2, 3, or 4 points for question 4, females receive 0 or 4 according to menstrual status.  
This difference creates smaller relative scores for females prior to menarche and larger relative scores after. 
67 
 
Significant linear increases across age were observed for both male (F(4, 66) = 23.67, p < 
.001) and female (F(4, 71) = 28.87, p < .001) participants. 
 
Agreement between Raters 
 
 
 Self- and mother-report pubertal scale scores were compared within gender.  For 
males, scores between self and mother were significantly correlated (r = .67, p < .001); as 
 
 
4.0
 
3.5  
 
3.0  
 
2.5
 
 
2.0
 
Gender
1.5  
Male
 
1.0 Female
 
10.00 11.00 12.00 13.00 14.00
 
Age
 
FIGURE 1.  Mean Puberty Category by Age and Gender. 
 
Mean Puberty Category
68 
 
were scores for females (r = .89, p < .001).  Paired samples t tests within gender were 
conducted to examine mean level differences between raters; male self-report scores (M = 
1.83, SD = .62) were significantly higher than were parent-report scores (M = 1.53, SD = 
.61; t(70) = -4.81, p < .001).  For females, self-report scores (M = 2.39, SD = .90) were 
significantly lower than parent-report scores (M = 2.54, SD = .87; t(74) = 3.07, p < .01). 
 
4.0  
 
3.5  
 
3.0
 
2.5  
 
2.0  
 
1.5
 Male
1.0  Female
10.00 11.00 12.00 13.00 14.00
 
Age  
Note:  Male and female pubertal status scores are calculated differently and cannot be 
compared across genders.  
 
Linear trend across ages: Male: p < .001; Femal e: p < .001. 
 
FIGURE 2.  Mean Pubertal Status Score by Age and Gender. 
Mean Pubertal Status Score
69 
 
 
Temperament and Puberty 
 
 
 Due to the uneven cell sizes in puberty category scores by gender, it was not 
possible to examine possible puberty category by gender interactions in temperament 
scores via the use of ANOVA.  Instead, pubertal status scores were standardized within 
gender and an interaction term was created for use in regression.  Significant puberty by 
gender interactions were found for self-report Surgency (β = .31, t = 2.03, p < .05) and, at 
the level of a trend, for parent-report Affiliation (β = .28, t = 1.85, p = .066) and Surgency 
(β = .30, t = 1.95, p = .053).  The poor range of pubertal scores in males made 
examination of changes across puberty in males somewhat more difficult and less reliable 
than similar examination in females; thus analyses were conducted by gender and are 
reported as such. 
 
Puberty Categories 
 
 
 A series of one-way ANOVA analyses with planned linear contrasts were 
performed within gender to determine possible differences in temperament scores as a 
function of pubertal stage scores.  For females, prepubertal participants were not included 
due to small cell size (n = 4); late pubertal males were similarly not included due to small 
cell size (n = 3).  No significant findings were revealed for males; however, female 
participants showed a decrease in Effortful Control across puberty at near trend level 
(F(2, 69) = 2.29, p = .11) with a significant linear trend (F(1, 69) = 4.59, p < .05) and is 
presented in Figure 3 as standardized (z) scores, rather than raw scores.  Examination of 
Effortful Control subscales revealed that Activation Control declined significantly (F(2, 
70 
 
69) = 7.16, p < .01, linear trend F(1, 69) = 12.60, p < .01) across puberty, and Inhibitory 
Control declined at near trend level (F(2, 69) = 2.21, p = .12, linear trend F(1, 69) = 3.26, 
p = .08).  Attention did not show a significant decline (See Figure 4 for illustration of 
standardized scores). 
 
 
 
.4  
 
 
 
.2  
 
 
 
-.0  
 
 
 
-.2  
 
 
 
-.4  
 
 
-.6  Effortful Control
 
Early Mid Late
 
Puberty Category  
 
One-way ANOVA  
p = .11; linear trend, p < .05  
 
 
 
 
FIGURE 3.  Female Standardized Effortful Control Scale Scores by Pubertal Status 
Group 
 
Mean Standardized Score
71 
 
 
 
.4
 
.2  
 
0.0
 
-.2  
 
-.4
 
-.6  
Activation Control
 
-.8 -_________________ .D  Attention -1.0 _____, Inhibitory Control
Early Mid  Late
Puberty Category  
 
 
One-way ANOVA 
Activation Control, p < .01; Inhibi tory Control, p = .12 
 
 
FIGURE 4.  Female Standardized Effortful Control Subscale Scores by Pubertal Status 
Group 
 
Mean Standardized Score
72 
 
 
.8
.6
.4
.2
-.0 -
-.2
-.4
-.6 Surgency
Early Mid Late
Puberty Category
 
 
One-way ANOVA, p < .01  
 
 
 
 
 
FIGURE 5.  Female Standardized Surgency Scale Scores by Pubertal Status Group 
 
Mean Standardized Score
73 
 
 Surgency scale scores increased significantly across puberty (F(2, 69) = 6.58, p < 
.01, linear trend F(1, 69) = 12.90, p < .01, see Figure 5 for illustration of standardized 
scores).  Examination of Surgency subscales revealed a significant decrease in Fear (F(2, 
69) = 15.70, p < .001, linear trend F(1, 69) = 30.28, p < .001), and an increase in High 
Intensity Pleasure at trend level (F(2, 69) = 2.76, p = .07, linear trend F(1, 69) = 4.47, p < 
.05).  Shyness did not show a significant decline (See Figure 6 for illustration of 
standardized scores).  Identical statistical tests were performed using mother-report 
temperament scores as the dependent variable; no significant results were found. 
 
Menarcheal Status 
 
 
 Female participants were coded as to menarcheal status.  Forty-nine participants 
had not reached menses; 27 had.  Independent samples t tests revealed that females who 
had reached menses reported themselves as significantly lower in Activation Control (M 
= 3.34, SD = .74) than girls who had not (M = 3.84, SD = .70, t(74) = 2.90, p< .01).  In 
addition, they reported lower scores on Fear (Post Menses:  M = 2.42, SD = .63; Pre-
Menses: M = 3.12, SD = .63, t(74) = 4.92, p < .001), and higher High Intensity Pleasure 
scores (Post Menses: M = 3.42, SD = .77; Pre-Menses: M = 2.98, SD = .77, t(74) = -2.37, 
p < .05).  Thus, overall Surgency scores were higher for girls who had reached menses 
(Post Menses: M = 3.56, SD = .57; Pre-Menses: M = 3.14, SD = .50, t(74) = -3.33, p < 
.01). 
 A significant difference also existed for self-reported Affiliation (see Figure 7).  
Females who had reached menses reported higher Affiliation scores than those who had 
not (Post Menses: M = 4.20, SD = .55; Pre-Menses: M = 3.85, SD = .64, t(74) = -2.44, p < 
74 
 
.01).  In addition, mother-report scores on Activation Control were significantly lower for 
females who had reached menses than for those who had not (Post Menses: M = 2.42, SD 
= .63; Pre-Menses: M=3.12, SD=.63, t(74)=4.92, p<.001). 
 
1.5
1.0
.5
• High Intensity 0.0 Pleasure
D Fear-Reverse Scored
-.5 --------------,..---~ . Shy-Reverse Scored
Early Mid Late
Puberty Category
 
 
One-way ANOVA  
Fear, p < .001; High Intensity Plea  sure, p = .07. 
 
 
 
FIGURE 6.  Female Standardized Surgency Subscale Scores by Pubertal Status Group 
 
Mean Standardized Score
75 
 
 .4
 
.3  
 
.2
 
.1  
 
-.0
 
-.1
 
-.2  
 
-.3 Affiliation
Pre Menarche P ost Menarche
Menarcheal Status Group  
 
Independent Samples t test; p < .01. 
 
 
FIGURE 7.  Female Standardized Affiliation Scale Scores by Menarcheal Status 
 
 
 
Pubertal Status Score, Age, and Temperament 
 
 
 Because age and puberty scores were highly correlated (age and puberty group: r 
= .62, p < .001, age and pubertal status score: r = .82, < .001), a series of zero order and 
partial correlations were conducted between temperament scores and the continuous 
Mean Standardized Score
76 
 
pubertal status score and age variables.  Self- and parent-report scores are summarized in 
Tables 7-8 (female) and 9-10 (male).  Results are summarized below by scale. 
 
 
 
TABLE 7.  Zero Order Pearson's Correlations and Partial Correlations 
Self-Report Temperament, Females 
Pubertal Puberty Age  Score Age Controlling Controlling for Age for Puberty 
Effortful Control -.27* -.10 -.26* .12 
Activation Control -.35** -.30** -.21+ -.07 
Attention -.07 .15 -.20+ .25* 
Inhibitory Control -.21+ -.03 -.24* .12 
Affiliativeness .10 .19+ -.07 .20+ 
Affiliation .24* .21+ .07 .12 
Perceptual Sensitivity -.08 -.03 -.08 .03 
Pleasure Sensitivity .09 .24* -.12 .26* 
Negative Affectivity .14 -.02 .16 -.12 
Aggression .16 .05 .13 -.05 
Depressive Mood .27* .10 .22+ -.08 
Frustration -.09 -.18 .02 -.15 
Surgency .38** .41** .11 .24* 
Fear -.44** -.47** -.21+ -.24* 
High Intensity Pleasure .37** .37** .13 .22+ 
Shyness -.06 -.06 .05 -.08 
+ p < .10, * p < .05, ** p < .01. 
 
77 
 
 
 
 
 
  
  
 TABLE 8.  Zero Order Pearson's Correlations and Partial Correlations 
 Parent-Report Temperament, Females 
  
 
 
 Pubertal 
Puberty Age 
 Group Age Controlling Controlling 
 for Age for Puberty 
 
 Affiliation -.01 -.12 .09 -.15 
 
 Effortful Control -.16 .10 -.24* .24* 
 Activation Control -.21+ .02 -.23* .17 
 
 Attention -.03 .14 -.13 .18 
 Inhibitory Control -.16 .10 -.26* .25* 
 
 Negative Affectivity .06 -.05 .16 -.11 
 
 Aggression .13 .05 .22+ -.04 
 Depressive Mood -.01 -.02 -.01 -.02 
 
 Frustration .02 -.15 .22+ -.21 
 
 Surgency .17 .17 .11 .08 
 Fear -.03 -.27* .19 -.30** 
 
 High Intensity Pleasure .20+ .16 .17 .05 
 Shyness -.12 .02 -.20+ .13 
 
 
 + p < .10, * p < .05, ** p < .01.  
 
78 
 
 
 
TABLE 9.  Zero Order Pearson's Correlations and Partial Correlations 
Self-Report Temperament, Males 
 
 
 Pubertal 
Puberty Age    
Group Age Controlling Controlling for for Age Puberty 
Effortful Control -.03 -.11 .06 -.12 
Activation Control -.13 -.23+ .04 -.19 
Attention .05 .05 .03 .02 
Inhibitory Control .04 -.03 .09 -.07 
Affiliativeness .14 .19+ .02 -.12 
Affiliation -.09 -.14 -.10 .04 
Perceptual Sensitivity .12 -.20 .36** -.39** 
Pleasure Sensitivity -.19 -.08 -.19 .08 
Negative Affectivity .19 .12 .15 -.02 
Aggression .14 .05 .08 .03 
Depressive Mood .23+ .12 .20+ -.05 
Frustration .04 .13 .05 -.03 
Surgency -.05 .00 -.07 .05 
Fear .11 -.06 -.01 -.04 
High Intensity Pleasure -.04 .01 -.06 .05 
Shyness .08 .04 .07 -.02 
+ p < .10, * p < .05, ** p <.01. 
 
79 
 
 
 
 
TABLE 10.  Zero Order Pearson's Correlations and Partial Correlations 
Parent-Report Temperament - Males 
 
Pubertal Puberty Age  Group Age Controlling Controlling for Age for Puberty 
Affiliation -.23* -.22+ -.11 -.08 
Effortful Control -.03 .01 -.05 .05 
Activation Control -.07 -.08 -.05 .01 
Attention -.03 .02 -.06 .05 
Inhibitory Control .05 .10 -.01 .07 
Negative Affectivity .19 .14 .15 -.02 
Aggression .09 .10 .02 .08 
Depressive Mood .09 .12 .02 .06 
Frustration .07 .13 -.04 .13 
Surgency -.14 -.04 -.15 .07 
Fear -.07 -.08 -.00 -.07 
High Intensity Pleasure -.06 -.01 -.08 .04 
Shyness .24* .12 .20+ -.04 
+ p < .10, * p < .05, ** p < .01. 
 
 
 
 
80 
 
Effortful Control 
 
 
For females, a significant negative relationship was found between Effortful 
Control and pubertal status score.  This relationship remained significant after controlling 
for the effects of age.  Activation Control was negatively related to both pubertal status 
and age, but only the relationship with pubertal status remained significant (at trend 
level) after controlling for age.  In addition, the negative relationship between Inhibitory 
Control and pubertal status strengthened from trend level to significance when 
controlling for age, and partial correlations revealed a negative relationship between 
Attention and pubertal status, along with a positive relationship between Attention and 
age. 
While parent-report of female Effortful Control showed no significant zero order 
relationship with either pubertal status or age, partial correlations revealed a negative 
relationship with pubertal status and a positive relationship with age.  A similar pattern of 
findings was present for Inhibitory Control, while Activation Control was negatively 
related to pubertal status at trend level (zero order correlations) and significantly so after 
controlling for puberty. 
For males, Activation Control was negatively related to age at trend level, but this 
relationship dropped to a less than significant level after controlling for pubertal status.  
There were no significant relationships between parent-report of male Effortful Control 
and puberty or age. 
81 
 
 
Affiliativeness 
 
 
In females, Affiliativeness was associated with age at trend level in both full and 
partial correlations, while the Affiliation subscale was associated with both pubertal 
status and age in full, but not partial, correlations.  Pleasure Sensitivity was associated 
with age across both types of correlations.  There were no significant relations between 
parent report Affiliation, age, and puberty for females. 
In males, self-report Affiliativeness showed a positive relationship at trend level 
with age, but not after controlling for puberty. Perceptual Sensitivity showed a positive 
partial correlation with age, and a negative partial correlation with puberty.  For male 
parent-report temperament, Affiliation was negatively related to both puberty and age, 
but neither relationship was significant after controlling for the other. 
 
Negative Affectivity 
 
 
 In females, self-report Depressive Mood showed a positive relationship with 
pubertal status that dropped to trend level when controlling for age; no other self-report 
Negative Affectivity subscale scores were related to pubertal status or age.  Parent-report 
Aggression and Depression in females were both positively related to pubertal status at 
trend level after controlling for age. 
For males, self-report Depressive Mood was positively related to pubertal status at 
trend level, both for zero order and partial correlations.  There were no significant 
relations between parent-reported Negative Affectivity in males and age or puberty. 
82 
 
 
Surgency 
 
 
In females, zero order correlations with Surgency were significant for both 
pubertal status and age; only the relationship with age remained significant after 
controlling for pubertal status.  Fear was negatively related to both pubertal status and 
age; both relationships remained significant after controlling for the other.  High Intensity 
Pleasure was positively related to both, but only the relationship with age remained 
significant after performing partial correlations. 
Parent-report of female Fear subscale was negatively related to age (both zero 
order and partial correlations), while High Intensity Pleasure was related to pubertal 
status at trend level, but only for the zero order correlation.  Shyness was negatively 
related at trend level to pubertal status, but only for the partial correlation. 
In male self-report, Shyness was positively related to puberty, but dropped to 
trend level after controlling for age.  No significant relationships were present for parent-
report Surgency scales or subscales for males. 
 
Temperament, Puberty and Socio-Emotional Behaviors 
 
 
 To determine possible roles of temperamental and pubertal processes in socio-
emotional behaviors, separate regression analyses were performed to predict Depressive 
Mood and Aggression.  In each regression, predictors were entered simultaneously.  
Effortful Control, Surgency, and Affiliativeness scale scores, as well as Frustration and 
Aggression subscale scores, were entered as temperament variables to predict Depressive 
Mood.  In addition, gender, age, and pubertal status score were also entered.  For 
83 
 
Aggression, predictors included Effortful Control, Surgency, and Affiliativeness scale 
scores, Frustration and Depressive Mood subscale scores, and gender, age, and pubertal 
status score. 
 The full model predicting Depressive Mood was significant (F(8, 137) = 12.53, p 
< .001, R2 = .42; see Table 11).  Examination of standardized Betas revealed that 
Effortful Control (β = -.38, t = -4.72, p < .001); Surgency (β = -.29, t = -4.715, p < .001); 
Pubertal Status (β = .27, t = 2.91, p < .01); Frustration (β = .18, t = 2.51, p < .05); and 
Affiliativeness (β = .14, t = 1.98, p = .05) all contributed significantly to prediction of 
Depressive Mood scores.  The direction of the relationships indicated that lower levels of 
Effortful Control and Surgency, higher Frustration and Affiliativeness, and more 
advanced pubertal status all predicted significant variance in Depressive Mood scores. 
  
 
TABLE 11.  Regression Analysis to Predict Depressive Mood 
Study 1 
 
 Beta t Sig. 
Gender .00 .02 .982 
Effortful Control -.38 -4.72 .000 
Affiliativeness .14 1.98 .050 
Surgency -.30 -4.27 .000 
Aggression .02 .20 .843 
Frustration .18 2.51 .013 
Pubertal Score .27 2.91 .004 
Age -.02 -.18 .861 
 
R2 = .42, p < .001 
 
84 
 
 Because age and pubertal status scores were highly correlated, the inclusion of 
both as predictors led to possible problems with collinearity and resulting instability of 
coefficients.  Thus, subjects were randomly split into two groups, and the regression was 
conducted separately for each group.  For Group 1, pubertal status again predicted 
Depressive Mood (β = .46, t = 3.07, p < .01), while age did not (β = -.14, t = -1.18, p = 
.24).  For Group 2, pubertal status was positively related to Depressive Mood, although 
not significantly so (β = .25, t = 1.45, p = .15), while age, again, was not (β = -.13, t =      
-86, p = .40). 
 The full model predicting Aggression was also significant (F(8, 137) = 11.96, p < 
.001, R2 = .41; see Table 12).  Examination of standardized Betas revealed that Effortful 
Control (β = -.47, t = -5.95, p < .001), Gender (β = -.22, t = -3.06, p < .01), and  
 
TABLE 12.  Regression Analysis to Predict Aggression 
Study 1 
 
Beta t Sig. 
Gender1 -.22 -3.06 .000 
Effortful Control -.47 -5.95 .000 
Affiliativeness -.05 -.67 .521 
Surgency .04 .53 .592 
Depressive Mood .02 .20 .843 
Frustration .18 2.36 .019 
Pubertal Score .09 .92 .357 
Age -.01 -.07 .946 
1Males are coded as 0, females as 1 
R2 = .41, p < .001. 
85 
 
Frustration (β = .18, t = 2.36, p < .05) all contributed significantly to the model.  The 
direction of the relationships indicated that lower levels of Effortful Control, higher 
levels of Frustration, and gender (male) all predicted significant variance in Aggression 
scores.  As neither age nor pubertal status predicted Aggression scores, additional 
analyses by groups were not conducted. 
 
Discussion 
 
 
 The primary goals of the current study were two-fold; first, to explore the 
hypothesis that emotional reactivity, as measured by the EATQ-R, would increase as a 
function of pubertal maturation; and second, to explore the relative contributions of 
temperament and pubertal maturation to aggression and depressive mood.  Study 1 also 
afforded an opportunity for further examination of psychometric properties of the EATQ-
R, particularly with regard to internal reliability, the relationship between scales, and 
agreement between raters.  Each of these goals will be discussed. 
 
Psychometric Properties of the EATQ-R 
 
 
In a previous study (Ellis & Rothbart, 2002) we found the EATQ-R to have good 
internal reliability, relationships between scales that were similar to those found in 
studies of child temperament, and modest agreement between self- and parent-report.  
One goal of the current study was replication of those findings. 
 In general, internal reliability of EATQ-R scales proved to be satisfactory in the 
current study.  Scale and subscale reliabilities were relatively comparable to those from 
the original study, with the exception of the Inhibitory Control subscale.  In the original 
86 
 
EATQ-R study, co-efficient alpha for Inhibitory Control was .65; in the current study, co-
efficient alpha was .56. 
The Inhibitory Control subscale contains only five items; Petersen et al., 
(Petersen, Crockett, Richards et al., 1988) assert that a five-item scale with reliability of 
.77 corresponds to an alpha of .88 for a tem-item scale.  Further, subscale reliability 
estimates in the current study are comparable to those obtained for self-report NEO-FFI 
(NEO-Five Factor Inventory) scales in a cross-cultural study of adolescent personality 
(McCrae et al., 2000).  Internal consistencies of 12 item NEO-FFI personality scales 
ranged from .57 - .86 in adolescents, while EATQ-R subscale reliabilities in the current 
study ranged from .52 - .79.  In addition, inter-item correlations for subscales averaged > 
.30.  Nonetheless, the alpha level is such that additional work may be warranted on the 
Inhibitory Control subscale.  Further, it is suggested that subscales not be separated from 
their “parent” scales in subsequent studies. 
 Agreement between reporters was also an issue of great interest in the current set 
of studies.  Parent report of temperament has been criticized by some as being biased and 
inaccurate (see Rothbart & Bates, 1998, for a complete discussion) but parents may, in 
fact, “be in a good position to observe the child’s behavior, especially infrequently 
occurring behavior that is nevertheless critical to defining a particular dimension of 
temperament” (Rothbart & Bates, 1998, p. 121).  Likewise, self-report may also be 
viewed as reflecting bias, particularly if responses tend to follow patterns of social 
acceptability.  However, to the extent that we can assume that adolescents respond with 
some degree of honesty when asked about their behavioral tendencies, their self-
perceptions are as important as the perceptions of other reporters, particularly with regard 
87 
 
to more internal processes such as emotions.  As such, parent- and self-report data were 
collected for each participant, allowing for comparison between reporters. 
Parent/self agreement for Study 1 was generally higher than would have been 
predicted by previous studies (Capaldi & Rothbart, 1992; Ellis & Rothbart, 2002).  In a 
1992 study utilizing the original EATQ, Capaldi and Rothbart reported agreement levels 
with genders combined that ranged from -.05 through .70, averaging .24.  In our previous 
study using the EATQ-R, we found agreement levels with genders combined of .05 
through.74, averaging .32 (Ellis & Rothbart, 2002).  Agreement in the current study 
ranged from .29 through .53 and averaged .40.  However, the mean age of the participants 
in the current study was younger than in the two previously cited studies.  Achenbach, 
McConaughy, and Howell (1987) report that agreement between reporters of behavioral 
and mental health problems, including parent/self agreement, is significantly higher for 6- 
to 11-year-olds than for older adolescents. 
 Examination of areas of agreement and disagreement between reporters revealed 
an interesting finding regarding Frustration.  Items in both self- and parent-report ask 
about irritability and frustration in the face of goal blocking.  While parent/self 
Frustration ratings in Study 1 showed moderate agreement (.30), adolescent participants 
consistently reported themselves higher in Frustration than did their parent.  This finding 
suggests that adolescent participants were not necessarily reporting their behaviors in a 
direction one might consider socially desirable.  However, it is important to note that 
Study 1 participants also consistently reported themselves as higher in Effortful Control 
and its subscales than did their parents. 
88 
 
Further differences were found upon examination of patterns of relationships 
between scales.  While parents’ reports of Surgency and Affiliation were significantly 
positively correlated (.27), this was not the case in self-report (.13), although the 
relationship was in the same direction.  This suggests that observable behaviors viewed as 
Affiliative may be quite different from Affiliative feelings and desires.  Parents may view 
extraverted behaviors as being Affiliative, thus linking such behaviors to Surgency, while 
adolescents may view Affiliative feelings as something different from extraverted 
behaviors.  In addition, parent-report Surgency in this sample was modestly positively 
related to Effortful Control and negatively related to Aggression.  Neither of these 
relationships was present for self-report. 
Overall, although there were some areas of disagreement between reporters, 
agreement was higher than had been expected.  However, areas of disagreement may 
provide as much information as do areas of agreement, as in the case of Affiliation.  
Achenbach et al. (1987) suggest that “each type of informant typically contributes a 
considerable amount of variance not accounted for by the others” (p. 227).  Jensen et al. 
(1999) further suggest that both parent and child informants are necessary to obtain 
accurate and adequate information for diagnosis of clinical conditions; this suggestion 
may be equally valid with regard to the measurement of adolescent temperament. 
 
Temperament, Puberty, and Age 
 
 
 It has been suggested that underlying neurological changes associated with the 
onset of puberty cause an increase in emotional reactivity that may lead to possible 
maladaptive outcomes (e.g., Dahl, 2001; Spear, 2000; Walker, 2002).  Others have 
89 
 
suggested that it is not puberty per se, but rather pubertal timing, which plays a role in 
psychosocial development (e.g., Graber, 1997; Stattin, 1990).  Indeed, there is empirical 
evidence to support both of these viewpoints (see Buchanan et al., 1992, for a review).  
However, the current study focused on pubertal maturation rather than pubertal timing 
because the high degree of interdependence between the two during early adolescence 
makes it difficult to assess independent contributions of each.  Further, the present study 
tested the hypothesis that physiological changes associated with puberty would affect 
physiologically based temperament systems.  Specifically, it was hypothesized that 
Negative Affectivity, Surgency, and in females, Affiliativeness, would increase across 
puberty, leading to increased emotionality. 
 To test for effects of pubertal maturation, one must also consider the role of age.  
While individuals differ in the timing and rate of pubertal development, there is still a 
strong, positive relationship with age.  Social transitions, such as the move from 
elementary- to middle-school are also closely tied to age; therefore, while the current 
study could not effectively control for social transitions, we did control for the effects of 
age.  Thus, each finding is discussed with regard to both age and pubertal maturation. 
The primary hypotheses were partially supported.  Levels of self-reported 
Surgency in females did, indeed, increase significantly across puberty, driven to some 
degree by an increase in High Intensity Pleasure, but to a greater degree by a decrease in 
Fear levels.  Partial correlations revealed the increase in High Intensity Pleasure to be 
more closely related to age than to pubertal maturation; however, the decrease in Fear 
was related to both age and pubertal maturation.  This suggests that, as females mature, 
they are less fearful and more apt to endorse items measuring high intensity pleasure.  
90 
 
Further, even when controlling for age differences in Fear levels, girls of advanced 
pubertal maturation were less fearful than were their less pubertally advanced age-mates.  
It is important to note, however, that parent-report female Fear was negatively associated 
with age, but not with pubertal maturation.  Further, there was no relationship between 
age, puberty, and self- or parent-report Fear in males. 
Levels of overall self- and parent-report Negative Affectivity did not increase 
with pubertal status; however, there was an increase in self-reported Depressive Mood 
with increased pubertal status in both males and females; this relationship remained 
significant even after controlling for the effects of age.  Interestingly, after controlling for 
age, parent-report female Frustration and Aggression were also positively related to 
pubertal maturation, as had been hypothesized.  This finding suggests that Negative 
Affectivity in adolescent girls may be more closely associated to pubertal maturation than 
to the changes that accompany age.  However, given the inconsistency of findings from 
previous studies with regard to puberty and negative affect, this finding needs to be 
replicated.  Further, the age range studied did not allow for sufficient variance in male 
pubertal status to fully test for a role of pubertal maturation in changes in temperament.  
An additional note of caution is warranted with regard to the partial correlations reported.  
Age and pubertal status were highly correlated; thus, the partial correlations may have 
been affected by issues of collinearity and instability of coefficients. 
Female Affiliativeness did not show a linear relationship with pubertal maturation 
as hypothesized; however, there was a significant increase in Affiliativeness with the 
onset of menarche.  This increase is consistent with the theory that affiliative behaviors 
may be related to pubertal activation of the neuropeptide oxytocin. 
91 
 
One of the most surprising findings in the current study was the decrease in both 
self- and parent-report female Effortful Control across pubertal maturation.  The decrease 
in Activation Control can be viewed as consistent with previous research showing a 
decrease in energy across pubertal maturation (see Buchanan et al., 1992, for a review), 
as most Activation Control items involve performing some sort of action.  However, 
decreases in Attention and Inhibitory Control were less expected. 
Our model of temperament is based in physiology; indeed, we believe that 
Effortful Control is related to functioning of frontal areas of the brain.  As reported in the 
introduction, laboratory studies suggest that performance on laboratory tasks purportedly 
involving frontal areas increases across adolescence (Davies & Rose, 1999; Levin et al., 
1991).  Further, neuro-imaging studies suggest that significant maturation of brain areas 
involved in executive functioning occurs during the adolescent period (e.g., Giedd, 1996; 
Klingberg, 1999; Paus, 1999; Sowell, 1998, 1999).  If Effortful Control is related to 
functioning of the executive system, we might expect it to increase during adolescence.  
In the current study, however, the opposite was true, at least with respect to puberty after 
controlling for age.  However, when controlling for the effects of puberty, parent-report 
Effortful Control increased with age as one might predict from studies of brain 
maturation. 
A concurrent decrease in Effortful Control and Fear, both of which are considered 
in our model to serve regulatory purposes, is both intriguing and perplexing.  However, it 
may be that increased physical maturation brings with it a decreased desire to present 
oneself as childlike and compliant.  Many of the Effortful Control items ask about 
behaviors, such as completing assignments and regulating impulsive behaviors, which 
92 
 
would be considered compliant.  Further, many of the Fear items are associated with fears 
and concerns that may be viewed as childlike by adolescents.  Thus, the pattern of 
responses may reflect changing attitudes toward compliance rather than changing 
temperaments.  Intuitively it seems reasonable to expect that such attitude change would 
also be associated with increasing age.  This was not the case in the current study. 
Alternatively, while the focus of the current study is on pubertal status, the 
findings may actually reflect the effects of pubertal timing.  By definition, many of the 
girls who are most pubertally advanced in the current study are also earlier maturing as 
compared to their age-mates.  It is not until the age of 14 or 15 that one can be reasonably 
sure that nearly all females are of equal pubertal status.  Thus, perhaps the current 
findings reflect attitudes toward compliance developed as a response to pressures and 
difficulties associated with early puberty.  The design of the current study does not allow 
exploration of this hypothesis. 
A third explanation may relate, in part, to Spears’ (2000) evolutionary theory of 
adolescent development.  Based on animal studies that show increased exploration of 
novelty during adolescence, Spears has suggested that changes in the adolescent brain 
important to frontal development, including increases in dopamine and decreases in 
glutamate and GABA in frontal areas, as well as changes in dopamine activity in limbic 
areas, may cause temporary alterations in the incentive value of stimuli such as drugs and 
social interactions.  Perhaps these brain changes in adolescence cause some sort of 
temporary perturbation of regulatory systems, resulting in increased dysregulation of 
behavior and emotion.  While such theories are intriguing, a great deal of empirical work 
remains to be done in the area of adolescent brain development before we can suggest 
93 
 
that general patterns of adolescent behavior reflect general patterns of adolescent brain 
development.  It is, however, a potentially exciting area of study. 
 
Temperament, Puberty, and Psychosocial Behaviors 
 
 
 The results of regression analyses to predict psychosocial behaviors in Study 1 
were generally consistent with expected results and with our previous study utilizing the 
EATQ-R (Ellis, 2002).  Low Effortful Control predicted the largest amount of variance in 
Depressive Mood scores, while low Surgency, high Affiliativeness, and Frustration, as 
well as increased pubertal maturation, also contributed.  Effortful Control may give an 
individual the ability to turn away from negative emotion, thus preventing the 
“rumination” often seen in depression (Nolen-Hoeksema & Girgus, 1994).  Alternatively, 
depression may erode an individual’s ability to control action and emotion.  The design 
of the current study does not allow us to determine directionality.  Likewise, the inverse 
relationship between Surgency and Depressive Mood may reflect a decrease in interest in 
activities and an increase in fear as symptoms of depression, or engagement in social 
activities may contribute to the onset of depression. 
Gender did not play a role in Depressive Mood in this sample, even though the 
majority of studies, including the original EATQ-R study, have found significant gender 
differences in depression.  However, most studies agree that gender differences in 
depression do not emerge until later adolescence (Nolen-Hoeksema & Girgus, 1994; 
Petersen, et al., 1988).  The relation between Depressive Mood and Affiliativeness was 
also present in the original EATQ-R study (Ellis, 2002).  Frank and Young (2000) 
94 
 
hypothesize that an increase in biologically driven affiliative needs in adolescent females 
results in increased salience of relationship quality.  Thus, relationship difficulties are 
more stressful for females than for males, resulting in increased depression in females.  
Indeed, empirical evidence suggests that relationship difficulties are associated with 
depression in adolescent females (Nolen-Hoeksema & Girgus, 1994).  We might expect, 
then, that the relationship between Affiliativeness and Depressive Mood would be 
mediated by relationship issues.  The current study did not measure such variables, and 
therefore cannot test for mediation effects.  However, Frank and Young’s theory (2000) 
would also account for the role of pubertal maturation in Depressive Mood. 
 Alternatively, the relationship between the two variables may reflect the role of 
loneliness in Depressive Mood.  Because the Affiliation scale asks about desires for 
affiliative experiences, rather than the actual presence of such experiences, high scores 
may indicate unfulfilled affiliative needs rather than the actual presence of a relatively 
high level of affiliative behaviors.  Future research on this issue will need to include 
relationship variables to understand fully the role of Affiliativeness in adolescent 
depression. 
 Effortful Control also showed a strong, negative relationship with Aggression, as 
found in a number of other studies (e.g., Eisenberg, 2001; Ellis, 2002; Rothbart, 1994).  
Frustration and gender (e.g., being male) also played a role.  In our previous study, 
Affiliativeness was negatively related to Aggression; this was not the case for this 
sample.  Further, Surgency was related to Aggression in our previous study; again, this 
was not the case in our current study.  However, the effect sizes of Surgency and 
Affiliativeness in the previous study were quite small compared to Effortful Control.  We 
95 
 
did not find a difference in gender in the previous study, but this may have been due to 
the older nature of the sample (mean age = 13.78 years), consistent with the finding that 
gender differences in aggression disappear by late adolescence (Finkelstein et al., 1994). 
 In sum, Study 1 provided at least partial support for our original hypotheses.  
While there was some indication of an increase in Depressive Mood with increasing 
pubertal maturation, intriguing evidence emerged that suggests that pubertal maturation 
may contribute to a decrease in both passive and active regulation systems.  However, the 
effect size of the relationship was modest and needs replication.  In addition, the study 
partially replicated our previous findings regarding the role of temperament in 
psychosocial behaviors.  Most notable was the contribution of poor Effortful Control to 
both Depressive Mood and Aggression, underscoring the importance of Effortful Control 
to healthy adolescent psychosocial development.  In addition, we replicated our previous 
finding regarding a relationship between Affiliativeness and Depression.  Finally, we 
obtained evidence of adequate internal reliability of the EATQ-R self- and parent-report 
forms in a group of relatively young participants.  In addition, agreement between 
reporters was generally higher than had been expected, suggesting that adolescent and 
parent participants in this age group tend to be in agreement with regard to general 
patterns of behavior in the adolescent. 
It is important to note that the adolescent participants in Study 1 represented a 
fairly homogenous group of late elementary- and early middle-school students in a 
primarily European-American middle-class community.  In this respect, the findings with 
regard to temperament and socio-emotional behaviors and the psychometric properties of 
96 
 
the EATQ-R could be expected to be similar to those obtained in our original study, as 
the participants were drawn from primarily the same demographic. 
 However, we felt it was important to replicate and extend our findings to an older, 
more diverse group of adolescents.  The majority of participants in Study 1 were young 
enough that it would be unlikely that many had started to participate in truly “anti-social” 
behaviors.  However, the study of Effortful Control in problem behaviors provided much 
of the impetus for revision of the EATQ-R.  Thus, it was important to study a population 
in which a certain percent were likely to be engaging in such behaviors.  Additionally, 
while risk and parenting variables were not assessed in Study 1, it was assumed that the 
majority of participants were from relatively low-risk environments; therefore, we could 
not examine the role of temperament in socio-emotional behaviors within the context of 
such variables. 
 An additional area of interest not addressed in Study 1 was the relationship, if 
any, between performance on executive attention measures and self- and parent-report 
Effortful Control.  As outlined in the introduction, our laboratory has found a positive 
relationship between performance on such tasks and Effortful Control in both toddlers 
(Gerardi-Caulton, 200X; Rothbart et al., 2002), and adults (D. Evans, personal 
communication, May 13, 2001).  However, because Study 1 was conducted entirely via 
the mail, it was not feasible to assess attention within the confines of such a study. 
We were very fortunate to have access to a subsample of participants currently 
involved in a longitudinal intervention study.  The participants are of high-school age and 
are diverse in terms of both ethnicity and risk variables.  Further, the participants are 
involved in repeated in-home and laboratory assessments of risk, parenting variables, and 
97 
 
both pro- and anti-social behaviors.  The level of assessment involved made inclusion of 
the attention tasks relatively simple.  Thus, access to Study 2 participants provided an 
ideal venue in which to answer the questions we hoped to address that went beyond those 
answered in Study 1.
98 
 
 
 
 
 
CHAPTER III 
 
 
STUDY II 
 
 
Introduction 
 
 
The specific goals of Study 2 were to explore relationships between performance 
on tasks assessing executive attention and temperament variables, as well as to assess the 
contributions of temperament, executive attention, and parenting and risk variables to 
psychosocial outcomes.  Specifically, it was hypothesized that performance on tasks 
assessing executive attention would be positively related to self- and parent-report 
Effortful Control and negatively related to self- and parent-report Negative Affectivity. 
Further, it was hypothesized that interference on an emotional Stroop task would 
specifically relate to depressive mood.  
In addition, it was hypothesized that temperament, attention, and parenting and 
risk variables would all contribute significantly to explanation of scores on psychosocial 
measures such as depressive mood, aggression, problem behaviors, and prosocial 
behaviors.  This chapter details the methods and results of Study 2, as well as a 
discussion of the findings. 
 
99 
 
 
 
Participants 
 
 
 Participants in Study 2 included 104 adolescents (63 female) and his or her parent 
or parents.  The participants represented a sub-sample from a larger longitudinal 
intervention study taking place in the Portland, Oregon, area.  The original sample was 
recruited from three area middle schools in one quadrant of the city representing a diverse 
community population in an area at risk for crime.  The ethnic makeup of the sub-sample 
used for the current project included 34 European Americans (11 female), 48 African 
Americans (34 female), and 22 members of other ethnic groups (16 female).  At the time 
participants were originally recruited, teacher ratings were used to identify low risk, at-
risk, and high-risk youth.  Risk was assessed using a 16-item measure developed to 
screen for risk in middle school (Teacher Risk Perception; adapted from Soberman, 1994, 
as ctied in Dishion, 1996).  The Teacher Risk Perception provides a brief, single-sheet 
instrument by which a teacher may quickly evaluation the risk status of all students in the 
class.  Areas of risk assessment include classroom behavior, tobacco use, involvement 
with troublesome or substance-using peers, and peer acceptance.  The current sample 
included 26 adolescents in the no-risk group (11 female), 35 in the at-risk group (19 
female), and 43 in the high-risk group (31 female).  All participants were 16- to 17-years-
of-age.  Parent participants included 44 fathers and 96 mothers. 
100 
 
Measures 
 
 
Questionnaire Measures 
 
 
Temperament Questionnaire 
 
 
 Adolescent and adult participants completed the EATQ-R adapted for use with 
scanning technology.  One reverse scored High Intensity Pleasure item (“I wouldn’t like 
to live in a really big city, even if it was safe.”) was deleted, because participants live in a 
large, metropolitan area. 
Child and Family Center Questionnaire – Child Form. 
 The Child and Family Center Questionnaire – Child Form is an adaptation of an 
instrument designed to measure parenting constructs and anti-social peer associations and 
behaviors (Metzler, Biglan, Ary, & Fuzhong, 1998).  Only youth report was utilized in 
the current study.  The parenting constructs assessed include parental monitoring of 
activities (e.g., “In the past three months, how often did at least one of your parents know 
what you were doing when you were away from home?”; 5 response options ranging 
from “Never or almost never” to “Always or almost always”); parental rule-making (“My 
parents _______ that I should do homework every day”; 4 response options ranging from 
“Didn’t have a rule or expectation” to “Had a clear rule”); positive family relationships 
(“Over the last month, I got along very well with my parents”; 5 response options ranging 
from “Never” to “Always”); and parent-child conflict (“In the last week did the following 
things happen between you and at least one of your parents?  We got angry at each 
other.”; 6 response options ranging from “Never” to “More than 7 times”).  The 
101 
 
questionnaire also contains scales designed to measure prosocial activities (“In the past 3 
months, how often did you participate in sport or organized activities?”; 5 response 
options ranging from “Never or almost never” to “Always or almost always” ); and 
prosocial peer affiliation (“In the last 3 months did your friends cooperate with their 
teachers?”; 5 response options ranging from “Never or almost never” to “Always or 
almost always”), as well as antisocial activities (“Please mark how many times you have 
done each of the following in the last month: Lied to your parents about where you have 
been or who you were with?”; 6 response options ranging from “Never” to “More than 20 
times”); and association with peers who engage in antisocial behaviors (“In the past 
week, how many times did you get together with friends who get into trouble a lot?”; 6 
response options ranging from “Never” to “More than 7 times”).  Appendix D contains a 
full listing of items by scale. 
A 1998 study utilizing the parenting scales found that youths’ reports of parenting 
constructs were stable over 3 quarterly waves of data collection (Metzler et al., 1998).  
Estimates of internal reliability ranged from .55 - .91.  Scores on parenting constructs 
were significantly correlated with youths’ reports of deviant behaviors and deviant peer 
associations. 
Scoring 
Mean scores were calculated for each scale, then standardized.  Standardized 
scores for prosocial activities and prosocial peer association were combined to form a 
prosocial composite score; antisocial activities and peer association were combined to 
form a problem behavior score. 
102 
 
 
Attention Measures 
 
 
General Procedures 
 
 
 The two attention tasks were presented on IBM compatible computers.  Some 
participants completed the tasks at home; others in an office used by Project Alliance 
staff.  Those completing the tasks at home used a laptop computer; those coming to the 
office used a desktop computer.  All participants completed the Attention Network Test 
first, followed by the Counting and Emotional Stroop Test. 
 
Attention Network Test 
 
 
 Participants completed the recently developed Attention Network Test (ANT; 
(Fan, McCandliss, Sommer, Raz, & Posner, 2002) designed to test the efficiency of the 
alerting, orienting, and executive attention systems.  These three attention systems are 
viewed to be distinct in both functional and anatomical terms (Posner & Peterson, 1990).  
The alerting system is involved in the achievement and maintenance of an alert state, the 
orienting system in the selection of information from sensory input, and the executive 
system in resolving conflict among responses.  The ANT produces reliable estimates of 
functioning of these three networks within a single task that can be performed by 
children, patients, and monkeys (Fan et al., 2002). 
 The executive attention system is often studied by conflict tasks, such as the 
Stroop color word task, which are known to activate areas of anterior cingulate and 
lateral prefrontal cortex.  An additional type of conflict task involves presentation of 
103 
 
stimuli surrounded by incongruent flankers.  Such tasks are also known to activate an 
area of anterior cingulate that overlaps, but is distinct from, areas activated by other 
conflict tasks.  The ANT utilizes a modification of a flanker task, and includes 
manipulation of warning cues that serve to, in some instances, activate the alerting 
system and, in others, to provide spatial information necessary for orienting.  A recent 
study utilizing the ANT with 40 adult subjects found significant test-retest reliability, 
with correlations between two sessions at .52 for alerting scores, .61 for orienting scores, 
and .77 for interference scores (Fan et al., 2002). 
 
Stimuli 
 
 
 Stimuli were presented on a computer screen.  Participants were presented a 
central fixation cross, then one of four warning cues: no-cue, center-cue, double-cue, or a 
spatial-cue.  The no-cue condition did not provide alerting or orienting cues.  The center-
cue consisted of an asterisk presented at the location of the fixation cross, and provided 
an alerting cue.  The double-cue consisted of two asterisks presented one degree above 
and one degree below the fixation cross.  This cue provided alerting but with a larger 
attentional field than in the center cue condition.  The spatial-cue consisted of an asterisk 
presented either one degree above or below the fixation cross, and indicated the location 
of the target stimuli to follow, thus providing an orienting cue. 
Participants were then presented target stimuli consisting of five arrows pointing 
left or right.  The target was a central arrow flanked on either side by two additional 
arrows identical in size and color to the central arrow.  In some trials, the flanking arrows 
pointed in the same direction as the central arrow.  These trials were considered 
104 
 
congruent trials.  In other, incongruent trials, the flanking arrows pointed in the opposite 
direction as the central arrow.  The participants were instructed to press a mouse key to 
identify the direction of the target arrow.  Figure 8 illustrates the experimental procedure. 
 
 
 
W a rning Cues 
 
 
    * 
 + + * * 
 * 
  
 No cue  Center Cue  Double Cue  Spatial  
 Cue 
 
Flanker Types 
 
           
           
 
Congruent Incongruent  
 
 
Events at Each Trial 
 
 
 fixation  I cue   I * ....ta.r.g.et  +  .. I 
 +  
100 0 +    + ms 100 ms 400  ms +  RT  <  
 500 1700  ms 
 ms 
 
 
 
FIGURE 8.  Attention Network Task Procedure 
 
 
 
105 
 
Procedure 
 
 
 Participants first read a set of instruction screens (see Appendix D for text of 
instructions).  Instructions indicated that a group of arrows would appear on the screen, 
and participants were instructed to press the button on the mouse that matched the 
direction that the center arrow was pointing.  Participants were told to maintain fixation 
on the center of the screen at all times. 
Each trial consisted of five events.  First, a central fixation cross was presented for 
a duration of 1000 ms.  This was followed on some trials by a warning cue presented for 
150 ms.  The fixation-cross then appeared for an additional 400 ms, followed by either 
congruent or incongruent target stimuli.  Target stimuli remained on the screen until the 
participant responded, at which point the computer registered the accuracy of the 
response as well as the reaction time in ms.  If the participant did not respond after 1700 
ms, the target disappeared and the computer registered a null response.  Each session 
consisted of 24 practice trials followed by two blocks of 64 trials each.  Within each 
block, half the trials were congruent, half were incongruent.  Each cue type occurred 
eight times.  Presentation was randomized within blocks. 
 
Scoring 
 
 
Three reaction time scores were calculated for each participant.  The first score, 
the interference score, was calculated by subtracting the mean RT of all congruent 
flanking conditions, summed across cue types, from the mean RT of incongruent flanking 
conditions.  For example, an individual with a mean RT of 650 ms for congruent trials 
106 
 
and a mean RT of 750 ms for incongruent trials would receive an interference score of 
100 ms.  Conversely, an individual with a mean RT of 650 ms for congruent trials but a 
mean RT of 800 ms for incongruent trials would receive an interference score of 150 ms.  
Errors were also calculated for each type of trial. 
The orienting effect was calculated as follows: The mean RT of spatial-cue 
conditions was subtracted from the mean RT of trials involving the center-cue.  The 
alerting effect was calculated by subtracting the mean RT of double-cued trials from the 
mean RT of no-cue trials.  Scoring procedures followed those developed by Fan et al. 
(2002). 
 
Emotion and Counting Stroop Task 
 
 
 Participants completed the Emotion and Counting Stroop Task, for which the 
methodology was adapted from two similar tasks designed for use in MRI imaging 
studies (Bush et al., 1998; Whalen et al., 1998).  Counting Stroop tasks involve 
presentation of number words that are either congruent or incongruent with the number of 
words presented on the screen.  Neuroimaging studies have found such tasks to activate 
an area of anterior cingulate near the area activated by conventional Stroop tasks.  The 
Emotion Stroop substitutes negatively valenced emotion words for neutral words in the 
task, and has been found in neuroimaging studies to activate an area of the anterior 
cingulate that is very near, but distinct from, the area activated by the Counting Stroop 
task.  The two tasks were combined to allow direct comparison of emotional and 
cognitive interference effects within subjects. 
 
107 
 
Stimuli 
 
 
 Stimuli were presented on a computer screen.  Participants were presented a 
central fixation cross, then a word presented in the middle of the screen, either one, two, 
three, or four times.  When words were presented multiple times, they were listed 
vertically in the middle of the screen.  Stimuli words were categorized as neutral, 
number, or emotion words, and are presented in Table 13.  Words were chosen from a 
corpus of words generated by children and adolescents for use in such tasks (Doost, 
Moradi, Taghavi, Yule, & Dalgleish, 1999). 
 
 
TABLE 13.  Emotion and Counting Stroop Stimuli 
 
Neutral Number Emotion 
Dog Sad 
Cat Cry 
Fox One Die 
Duck Two Fear 
Horse Three Worry 
Whale Four Crazy 
Sheep Panic 
Rabbit Lonely 
 
 
 
 
Procedure 
 
 
 Participants first read a set of instruction screens (see Appendix E).  Instructions 
indicated that a group of 1-4 words would appear on the screen, and participants were 
108 
 
instructed to press a button on the keyboard that corresponded with the number of words 
presented on the screen. 
Each trial began with a central fixation cross, presented for a duration of 1000 ms.  
This was followed by presentation of target stimuli which remained on the screen until 
the participant responded, at which point the computer registered both the accuracy of the 
response as well as the reaction time in milliseconds.  If the participant did not respond 
after 1700 ms, the target disappeared and the computer registered a null response. 
Each session consisted of 24 practice trials followed by one block of 96 trials 
each; 32 neutral, 32 number, and 32 emotional trials.  Each neutral cue, as well as each 
emotion cue, was presented four times, with each word presented in groups of 1, 2, 3, and 
4 words.  Each number trial was presented eight times in various combinations of 1, 2, 3, 
or 4 words per trial.  However, every number trial was incongruent; that is, the words 
used for the stimuli never matched the number of times it was presented.  Presentation 
was randomized. 
 
Scoring 
 
 
Two interference scores were calculated for each participant.  The first, the 
Counting Stroop interference score, was calculated by subtracting the mean RT for 
neutral trials from the mean RT for number trials.  The second, the Emotion Stroop 
interference score, was calculated by subtracting the mean RT for neutral trials from the 
mean RT for emotion trials.  Errors for each trial were also calculated. 
109 
 
RESULTS 
 
 
Questionnaire Measures 
 
 
EATQ-R 
 
 
Internal Consistency 
 
 
A series of item-total analyses were performed by reporter to estimate coefficient 
alpha (Cronbach’s) for each scale and subscale.  Results are reported in Table 14.  For 
self-report scales, alpha ranged from .70 - .81 and alpha for subscales ranged from .43 - 
.82 (see Table 14).  For parent-report, scale alphas ranged from .69 - .81 and alpha for 
subscales ranged from .50 - .84.  Average inter-item correlation was calculated for each 
subscale to further insure internal consistency even within small clusters of items 
covering broad dimensions of behavior.  Subscale average inter-item correlations 
averaged .35.  
 
Agreement between Reporters 
 
 
 Agreement between self- and mother-report by gender is reported in Table 15, as 
well as agreement between self- and father-report, between self- and combined mother-
father report scores, and between mother- and father-report.  Agreement between female 
participants and mothers ranged from -.07 to .48, and was statistically significant for the 
Attention, High Intensity Pleasure, Fear, Shyness, and Depressive Mood subscales, as 
well as overall Surgency.  In addition, agreement on the Affiliation subscale was 
significant at the level of a trend.  For males, correlation with mother-report ranged from 
110 
 
-.07 to .40 and reached significance for subscales measuring Shyness and Aggression 
and, at trend level, Inhibitory Control. 
 Too few fathers completed the EATQ-R to allow examination of agreement by 
gender; however, agreement with genders combined ranged from -.09 to .41, and reached 
significance for subscales measuring Inhibitory Control, High Intensity Pleasure, 
Depressive Mood, and Aggression, and, at the level of a trend, overall Surgency.  
Combining mother and father report yielded levels of agreement ranging from -.01 to .51, 
and reached significance for the Inhibitory Control, High Intensity Pleasure, Depressive 
Mood, and Aggression subscales and, at trend level, the Attention and Shyness subscales.  
Further, agreement was statistically significant for overall Surgency and, at trend level, 
overall Negative Affectivity.  Combining parent reports improved agreement on 
Surgency and Negative Affectivity above that achieved by either parent alone; however, 
agreement on Effortful Control was slightly poorer than for father-report alone.2  Mother- 
and father-report agreement ranged from .31 to .75, and was statistically significant for 
all scales and subscales. 
 
Mean Differences between Reporters 
 
 
 Scale means by gender, and reporter (self and mother) are presented in Table 16.   
 Female participants rated themselves higher in Affiliation (t(55) = 2.10, p < .05) and 
Frustration (t(55) = 2.05, p < .05) than did their mothers, and lower in Aggression (t(55) 
= -1.79, p = .08; overall Negative Affectivity; self-report higher; (t(55) = 1.72, p = .08) at 
                                                 
2 Mother report for those individuals whose fathers completed questionnaires were similar to levels of 
agreement for all mothers in Sample 2.  However, agreement on the Inhibitory Control subscale was .41. 
111 
 
trend level.  Males rated themselves significantly higher in Attention (t(39) = 2.31, p < 
.05) than did mothers, and lower in Aggression (t(39) = -1.99, p = .054) at trend level. 
 
TABLE 14.  Cronbach’s Alpha Scale and Subscale Reliabilities by Reporter 
Study 2 
 
Scale Subscale Self-Report Mother-Reportn = 106 N = 90 
Effortful Control  .81 .78 
 Activation Control .64 .79 
 Attention .55 .72 
 Inhibitory Control .43 .50 
Affiliativeness  .81 - 
 Affiliation .65 .72a 
 Perceptual Sensitivity .63 - 
 Pleasure Sensitivity .72 - 
Surgency  .70 .69 
 High Intensity Pleasure .53 .68 
 Fear (reverse scored) .61 .67 
 Shyness (reverse scored) .75 .84 
Negative Affect  .83 .81 
 Frustration .82 .76 
 Depressive Mood .78 .74 
 Aggression .69 .79 
aOnly the Affiliation subscale is included in parent report Affiliativeness 
112 
 
 
 
 
TABLE 15.  Temperament Score Correlations between Reporters 
Sample 2 
 
Self- Self-
 Mother, Mother, Self-
Self-
Mother/Father Mother- 
Female Male Father Combined Father 
n = 51 n = 39 n = 40 n = 35 n = 35 
Effortful Control .10 .12 .20 .18 .75** 
Activation   
Control .17 .03 .22 .12 .50** 
Attention .28* .26 .19 .29+ .73** 
Inhibitory 
Control .04 .29+ .40* .38* .69** 
Affiliation .30+ .01 .09 .25+ .31+ 
Surgency .48** .09 .26+ .38* .60** 
High Intensity 
Pleasure .31* .23 .41** .51** .53** 
Fear  .28* -.07 .21 .21 .66** 
Shyness  .27* .39* .24 .29+ .51** 
Negative Affect .13 .01 .22 .25+ .64** 
Frustration -.07 .17 -.09 -.01 .40* 
Depressive 
Mood .45** .04 .39** .47** .74** 
Aggression .18 .32* .30* .36* .46** 
 
+ p < .10, * p < .05, ** p < .01. 
113 
 
 
 
TABLE 16.  Temperament Scores by Gender and Reporter, Sample 2 
 
Self,  Mother, 
Female Self,  Male 
Mother, 
Female Male 
Scale 
M (SD) M (SD) 
M (SD) M (SD) 
n = 40 
n = 56 n = 51 n = 39 
Effortful Control 3.44(.50) 3.35(.47) 3.34(.47) 3.18(.57) 
Activation   
3.06(.70) 3.06(.60) 2.94(.76) 2.77(.79) 
Control 
Attention 3.51(.65) 3.41(.60) 3.50(.54) 3.25(.77)* 
Inhibitory Control 3.70(.59) 3.56(.57) 3.62(.59) 3.62(.61) 
Affiliation 3.82(.68)  3.62(.72) * 3.58(.72) 3.46(.71) 
Surgency 3.37(.57)  3.44(.45)  3.63(.49) 3.66(.50) 
High Intensity 
3.14(.76) 3.20(.55) 3.27(.59) 3.31(.66) 
Pleasure 
Fear  2.56(.66)  2.41(.66)  2.18(.62) 2.04(.71) 
Shyness  2.43(.98) 2.46(.83) 2.29(.95) 2.30(.78) 
Negative Affect 2.70(.58) 2.63(.50) + 2.48(.56) 2.59(.57) 
Frustration 3.23(.85) 2.92(.69) * 3.04(.81) 3.09(.75) 
Depressive Mood 2.68(.79)  2.51(.81)  2.09(.69) 2.09(.69) 
Aggression 2.23(.69) 2.44(.66)+ 2.26(.73) 2.53(.76)+  
Paired Samples t test 
Difference between parent- and self-report: + p < .10, * p < .05, ** p < .01. 
 
 
114 
 
 
 
Correlations between Scales 
 
 
 Correlations between temperament scale scores and scores on the two socio-
emotional scales are presented in Table 17.  Both Aggression and Depressive Mood were 
negatively correlated with Effortful Control and positively correlated with Frustration, 
and with each other, in both parent- and self-report.  In addition, Depressive Mood was 
negatively related to Surgency in both parent- and self-report.  However, in self-report 
Affiliativeness showed a modest, but significant, positive correlation with Depressive 
Mood and a modest negative correlation with Aggression.  For parent-report, only the 
negative relationship with Aggression was evident. 
 
Gender Differences 
 
 
 Females in Study 2 reported themselves as significantly higher on subscales 
measuring Depressive Mood (t(101) = -3.95, p < .001), Pleasure Sensitivity (t(101) =       
-3.73, p < .001), and Fear (t(101) = -3.03, p < .01) than males in their sample, and higher 
in  both Affiliation (t(101) = -1.91, p = .06) and Perceptual Sensitivity (t(101) = -1.91, p 
= .054) at trend level.  Gender differences in scale scores reached statistical significance 
for Affiliativeness (females higher, t(101) = -3.19, p < .001), and Surgency (males higher, 
t(101) = 2.32, p < .001), and at trend level for Negative Affectivity (females higher, 
t(101) = 1.99, p = .05). 
 
115 
 
TABLE 17.  Pearson’s Correlations between Scales, Study 2 
Pearson’s Correlations – Self-Report Temperament 
 
 1 2 3 4 5 6 
1-Effortful Control - .26** .11 -.58** -.34** -.36**
2-Affiliation .11 - .32** -.15 .04 -.06 
3-Surgency .07 -.08 - -.00 -.39** -.03 
4-Aggression -.42** -.07 -.08 - .17+ .56**
5-Depressive Mood -.36** .24** -.46** .31** - .15 
6-Frustration -.33** .25** -.20* .29** .43** - 
Parent-report values above the diagonal, self-report below 
+ p < .10, *p < .05, **p < .01 
 
Child and Family Center Questionnaire – Child Form (CFCQC) 
 
 
 Forty-three males and 60 females completed the CFCQC.  Scale scores are 
summarized in Table 18.  Independent samples t tests revealed a significant gender 
difference in Deviant Peer Association scores (t(101) = 2.07, p < .05), such that males 
reported greater association with deviant peers than did females.  Conversely, females 
reported greater association with prosocial peers than did males, but only at trend level 
(t(101) = -1.78, p = .08).  No other significant gender differences were observed. 
 Correlations between CFCQC scale scores are presented in Table 19.  As 
illustrated, Deviant Peer Relation and Deviant Behavior Scores were highly correlated (r 
= .78, p < .001).  In addition, Parental Monitoring, Relationship Quality, Parental Rules, 
and Prosocial Peer association were all significantly negatively related to both Deviant 
Behavior and Deviant Peer Association (see Table 19 for values).  Further, Family 
116 
 
TABLE 18.  CFCQC Scale Scores by Gender 
 
Scale Female Male Combined M (SD) M (SD) M(SD) 
Parental Monitoring 2.71 (1.11) 2.72 (.97) 2.71 (1.05) 
Parental Rules 1.99 (.72) 1.90 (.67) 1.95 (.72) 
Family Conflict .60 (.77) .71 (.64) .65 (.72) 
Relationship Quality 2.45 (1.02) 2.49 (.92) 2.47 (.98) 
Deviant Peer Associations .58 (.67) .87 (.71)* .70 (.69) 
Deviant Behaviors .34 (.48) .35(.44) .34 (.46) 
Pro-social Peer Associations 2.56 (.70)+ 2.31 (.70) 2.43 (.71) 
Pro-social Behaviors 2.57 (.60) 2.51 (.70) 2.54 (.64) 
Independent Samples t test;  
Significant Gender Difference: + p < .10, * p < .05 
 
Conflict was positively related to Deviant Peer Relations and, at trend level, to Deviant 
Behavior.  Prosocial Peers and Prosocial Behavior were positively related, and were both 
related to Parental Monitoring, Relationship Quality, and Parental Rules.  In addition, 
Prosocial Peer scores were negatively related to both Deviant Behaviors and Deviant 
Peers (see Table 19). 
 
Attention Measures 
 
 
Attention Network Test 
 
 
 Forty-one males and 61 females completed the ANT; however, data for 1 male 
and 4 females were not included in the analysis because their scores indicated a pattern of 
responding to flanker direction rather than target direction.  Median reaction time data by 
117 
 
flanker and warning cue type were calculated for remaining participants.  Means of these 
calculations are presented in Table 20.  A 2 x 4x 2 (Flanker x Warning Type x Gender) 
repeated measures analysis of variance revealed a significant main effect of flanker type 
(F(1, 98) = 378.35, p < .001), such that Reaction Times (RTs) for trials involving 
incongruent flankers were longer than those involving congruent flankers.  In addition, 
there was a significant main effect of warning cue type (F(3, 294) = 57.68, p < .001), and 
planned linear contrasts revealed that warning types significantly decreased across types 
(F(1, 98) = 137.06, p < .001; no-cue, center-cue, double-cue, location-cue) as illustrated 
in Figure 9.  Additionally, tests of between subjects effects revealed a significant main 
effect of gender F(1, 98) = 6.97, p < .05), such that males had significantly shorter RTs 
than did females (see Figure 10). 
 
TABLE 19.  CFCQC Correlations between Scales 
 
Scale 1 2 3 4 5 6 7 
1-Parental 
Monitoring -       
2-Parental Rules .35** -      
3-Family Conflict -.04 -.13 -     
4-Relationship 
Quality .42** .20* -.47** -    
5-Deviant Peer 
Associations -.34** -.34** .30** -.38** -   
6-Deviant 
Behaviors -.42** -.24* .19+ -.40** .78** -  
7-Pro-social Peer 
Associations .43** .33** -.01 .26** -.23* -.22* - 
8-Pro-social 
Behaviors .49** .32** -.03 .41** -.12 -.15 .58**
+ p < .10; * p < .05; ** p < .01. 
 
118 
 
 
800  
 
 
 
 
 
700  
 
 
 
 
 
600  
 
 Flanker 
 
 ■ Congruent
 
500  * Incongruent
None Center  Double Location
 
Warning Cue  
 
Main effect of  flanker type; p<.001 
Main effect of  warning cue type; p<.001 
 
 
FIGURE 9.  ANT RTs by Flanker and Warning Cue Type 
 
 
 
 
 Errors were also submitted to a 2 x 4 x 2 (Flanker x Warning Type x Gender) 
repeated measures analysis of variance, revealing a significant main effect of flanker 
(F(1, 98) = 29.78, p < .001), indicating that participants made significantly more errors 
during trials involving incongruent trials (see Figure 11).  No other main effects or 
interactions were present. 
Mean RTs in milliseconds
119 
 
 
 A series of RT subtractions were performed, and results are presented in Table 21, 
as well as error rates by trial type and gender, and overall RTs.  Alerting scores were 
calculated as the difference between no cue and double cue trials; orienting as the 
difference between center cue and location cue trials; and interference as the difference 
between congruent and incongruent trials.  There were no significant gender differences 
between RT subtraction scores. 
 
 800
 
 
 700
 
 
 
600
 
Gender
 
Male
 
500 • Female
 Congruent Incongruent
 Flanker Type
Gender; p < .05 
 
FIGURE 10.  ANT RTs by Gender and Flanker Type 
Mean RTs in milliseconds
120 
 
 
 
 
 
 
.6  
 
 
.5  
 
 
 
.4  
 
 
 
.3  
 
 
.2  
 
 Flanker
 
.1  
 ■ Congruent
 
0.0  * Incongruent
None Center  Double Location
 
 
Warning Cue  
 
 Main effect of flanker type; p < .001 
 
 
 
 
FIGURE 11.  ANT Errors by Flanker and Warning Cue Type 
 
 
 
Mean Errors
121 
 
Table 20.  ANT RTs by Gender, Flanker Type, 
and Warning Cue Type 
 
Warning Type 
Gender Flanker Type 
No Cue Center Cue Double Cue Spatial Cue 
Male Congruent 
599.61 578.17 561.86 546.62 
(88.80) (83.40) (77.76) (75.47) 
n = 40 Incongruent 694.13 702.74 692.30 655.04 (89.51) (113.97) (124.97) (119.54) 
644.90 630.55 609.41 598.26 
Female Congruent (86.11) (91.36) (84.64) (95.09) 
n = 61 
Incongruent 779.45 759.87 744.23 714.15 (127.09) (127.01) (119.46) (129.84) 
Congruent 630.47 611.29 592.32 580.69 
Total (97.67) (89.35) (89.35) (98.65) 
Incongruent 744.41 734.16 720.87 687.55 (122.13) (124.04) (124.09) (128.14) 
 
 
 
 
TABLE 21.  ANT Attentional Network Subtractions by Gender 
 
 
 Attentional Networks Subtractions   Errors Errors 
Gender Orienting Conflict Overall RT Congruent IncongruentAlerting 
19.70 47.97 107.49 599.09 .23 1.48 
Male 
(33.08) (41.83) (43.44) (77.41) (.68) (1.92) 
24.07 35.19 118.61 653.35 .58 1.40 
Female 
(51.62) (38.87) (56.56) (87.39) (1.24) (2.01) 
22.10 40.93 113.62 628.99 .43 1.44 
Total 
(44.11) (40.52) (51.12) (87.98) (1.04) (1.96) 
 
122 
 
Counting and Emotional Stroop Task 
 
 
 Forty-four males and 61 females completed the Counting and Emotional Stroop 
Task.  However, one male participant had errors in excess of 4 standard deviations above 
the mean; thus his scores were eliminated from further analyses.  RT, conflict, and error 
rates for remaining participants by gender and trial type are summarized in Table 22.  A 3 
x 2 (Trial Type x Gender) repeated measures analysis of variance revealed a significant 
main effect of type, (F(2, 202) = 28.91, p < .001), and planned simple contrasts revealed  
 
 
TABLE 22.  Counting and Emotional Stroop RTs, Conflict, and Errors 
By Gender and Stimulus Type 
 
Gender  Stimulus Type  Neutral Number Emotion Total 
Reaction Time 676.51 704.26 677.07 690.50 (117.64) (131.38) (119.11) (136.19) 
Male 
n = 44 Conflict Effect N/A 
38.99 5.30 22.14 
(71.61) (40.29) (48.08) 
Errors .47 1.77 .67 2.91 (.88) (1.36) (.81) (1.87) 
Reaction Time 769.02 812.30 767.93 783.08 (143.32) (170.17) (151.00) (151.16) 
Female 43.28 -1.09 21.10 
n = 61 Conflict Effect N/A (73.58) (38.09) (48.60) 
Errors .30 1.02 .44 1.75 (.64) (1.25) (.81) (2.01) 
 Reaction Time 731.64 773.13 733.22 746.00 (142.17) (171.71) (149.59) (150.97) 
Total Conflict Effect N/A 41.49 1.58 21.53 (72.60) (38.96) (48.15) 
 Errors .37 1.33 .54 2.24 (.75) (1.34) (.82) (2.03) 
 
123 
 
TABLE 23.  Pearson’s Correlations – ANT Measures 
Genders Combined 
 
 
 1 2 3 4 5 
1-Alert -     
2-Orient .14 -    
3-Interference .01 -.03 -   
4-Congruent Errors -.11 -.16 -.24* -  
5-Incongruent Errors -.12 -.24* .16 .30** - 
6-Mean RT -.12 -.07 .46** .18+ -.12 
*p < .05, ** p < .01. 
 
 
 
that the difference existed between neutral and number trials (F(1, 101) = 33.07, p < 
.001), while there was no significant difference between neutral and emotion trials.  In 
addition, there was a significant main effect of gender (F(1, 101) = 9.39, p < .001), such 
that females had longer reaction times than did males (see Figure 12).  While emotion 
trials did not produce significantly higher RT than neutral trials, an examination of 
distributions showed that many participants had shorter RTs during emotion than neutral 
trials, thus this measure was retained as a measure of interference. 
 Errors were also submitted to a 3 x 2 (Type x Gender) repeated measures analysis 
of variance.  There was a significant main effect of type, (F(2, 202) = 35.76, p < .001), 
and planned simple contrasts revealed that the difference existed between neutral and 
number trials (F(1, 101) = 48.03, p < .001).  In addition, errors for emotion trials were 
higher than errors for neutral trials, but only at trend level (F(1, 101) = 3.46, p = .06).  A 
significant main effect of gender was present (F(1, 101) = 8.43, p < .01), indicating that 
124 
 
 820
 800
 
780
 
760
 
740
 
 720
 700 Gender
 680 T Male
 660 • Female
Neutral Emotion Number
 
 StimulusType
 Main effect of stimulus type; p < .001 
Main effect of gender; p < .001 
 
FIGURE 12.  Counting and Emotional Stroop RTs by Stimulus Type and Gender 
 
 
males made significantly more errors overall than did females.  In addition, a Type x 
Gender interaction was present (F(2, 202) = 3.20, p < .05), such that the gender 
difference occurred primarily in number trials (see Figure 13). 
 Stroop number conflict effects were calculated as a difference between neutral 
and number trials; emotion conflict effects were calculated as the difference between 
neutral and emotion trials.  Independent samples t test revealed no significant gender 
differences in conflict effects. 
Mean RTs in milliseconds
125 
 
 
 
 2.0
 
 
1.5
 
 
 1.0
 / / ,,,,..,,,,.. 
 
.5 ___________ / / .,,,, _,,,.. Gender
 
Male
 
0.0 Female
 Neutral Emotion Number
Stimulus Type   
 
Main effect  of stimulus type; p < .001 
Main effect of gender; p < .01 
 
 
 
FIGURE 13.  Counting and Emotion Stroop Errors 
By Stimulus Type and Gender 
 
Mean Errors
126 
 
Correlations between Attention Measures 
 
 
 Correlations between various aspects of performance on the attention tasks are 
presented in Tables 23-29.  With genders combined, the three attention network scores 
measured by the ANT were uncorrelated.  As gender differences were noted in 
performance, however, correlations were also examined by gender.  In females, the 
alerting and orienting effects were modestly related (r = .25, p = .06).  Errors were related 
in females (r = .42, p < .01), but not in males.  In the Counting and Emotion Stroop task, 
Emotion and Counting Stroop Interference were related in the full sample (r = .47, p < 
.01) as well as in both genders.  Further, errors across trial types were related in the full 
sample, but to a greater degree in males than in females.  Errors across the two tasks were 
related in the full sample, but primarily for females.  RTs for the two tasks were also 
related. 
 
 
TABLE 24.  Pearson’s Correlations – ANT Measures 
by Gender 
 
 1 2 3 4 5 6 
1-Alert - -.02 -.07 .08 .06 -.28+ 
2-Orient .25+ - -.02 -.18 .08 .02 
3-Interference .08 .04 - -.31 .15 .44**
4-Congruent Errors -.17 -.29* -.32* - .06 -.12 
5-Incongruent Errors -.21 -.43** .17 .42** - -.40**
6-Mean RT -.08 -.07 .47** .17 .06 - 
Values for male participants fall above the diagonal; values for females fall below     
* p < .05, ** p < .01. 
 
127 
 
 
 
TABLE 25.  Pearson’s Correlations – Counting and Emotion Stroop Measures 
Genders Combined 
 
 
 1 2 3 4 5 
1-Stroop Interference -     
2-Emotion Interference .47** -    
3-Neutral Errors .06 .16 -   
4-Number Errors -.01 -.06 .07 -  
5-Emotion Errors -.03 -.01 .25* .30** - 
6-Mean RT .33** .21* -.07 -.09 -.29**
* p < .05, ** p < .01 
 
 
 
 
TABLE 26.  Pearson’s Correlations – Counting and Emotional Stroop 
Measures by Gender 
 
 
 1 2 3 4 5 6 
1-Stroop Interference - .43** -.14 -.10 -.10 .27+
2-Emotion Interference .50** - -.14 -.29+ -.15 .13 
3-Neutral Errors .25* .42** - -.15 .22 -.27+
4-Number Errors .08 .07 .24+ - .15 -.03 
5-Emotion Errors .04 .07 .26* .37** - -.20 
6-Mean RT .34** .29* .15 .00 -.30* - 
 
Values for male participants fall above the diagonal; values for females fall below        
* p < .05, ** p < .01. 
 
128 
 
TABLE 27.  Pearson’s Correlations between Attention Task Measures 
Genders Combined 
 
 
 Stroop  Emotion  Neutral Number Emotion Mean
Inter. Inter. Errors Errors Errors RT 
Alert -.01 -.23* -.12 -.05 .03 -.19
Orient .01 .12 -.02 -.11 .03 -.09
Interference .00 -.09 -.09 -.07 -.10 .09
Congruent Errors .02 .06 -.04 -.04 .03 .23*
Incongruent Errors -.17+ -.01 .02 .31** .26** -.12
Mean RT - ANT .22* .16 -.09 -.27** -.34** .60**
* p < .05, ** p < .01 
 
 
TABLE 28.  Pearson’s Correlations between Attention Task Measures 
Females 
 
 Stroop  Emotion  Neutral Number Emotion Mean
Inter. Inter. Errors Errors Errors RT 
Alert -.16 -.28* -.24+ -.12 -.03 -.21
Orient .01 .08 .14 -.17 -.02 -.22
Interference -.04 -.11 -.12 -.00 -.02 .09
Congruent Errors -.01 .10 .00 .01 .05 .23+
Incongruent Errors -.04 .15 -.20 .33* .39** .03
Mean RT - ANT .16 .20 -.04 -.19 -.28* .49**
* p < .05, ** p < .01 
 
 
129 
 
TABLE 29.  Pearson’s Correlations between Attention Task Measures 
Male 
 
 Stroop Emotion Neutral Number Emotion  Mean 
Inter. Inter. Errors  Errors Errors RT 
Alert .32* -.12 .02 .11 .16 -.24 
Orient .03 .16 -.17 -.15 .05 .20 
Interference .06 -.04 -.07 -.10 -.17 .01 
Congruent Errors .09 -.08 -.13 .11 .25 -.13 
Incongruent Errors -.36* -.26+ .24 .29+ .10 -.40**
Mean RT - ANT .30+ .14 -.07 -.21 -.34* .69**
* p < .05, ** p < .01 
 
 
Adjusted Interference Scores 
 
 Examination of RTs and errors revealed that, in males, overall ANT reaction time 
was negatively associated with errors.  This pattern indicates that participants may have 
been sacrificing accuracy for speed.  Thus, an individual may have low interference 
scores, but a high number of errors.  This same pattern was also evident for females in 
Emotion errors on the Counting and Emotion Stroop, and in Number errors for males. 
To examine the effects of this speed-accuracy tradeoff more closely, scores for 
interference and errors were standardized within each task and recoded as follows for 
each task: 1= Low interference and errors (> 1 SD below the mean on interference, 
combined with below average errors, or > 1 SD below the mean on errors, combined with 
below average interference); 2 = Average interference and errors (both interference and 
130 
 
errors fall within 1 SD of the mean); 3 = Moderate interference and errors (> 1 SD above 
the mean on interference, combined with above average errors, or > 1 SD above the mean 
on errors, combined with above average interference); and 4 = High interference and 
errors (> 1 SD above the mean on both interference and errors).  Individuals scoring > 1 
SD below the mean on interference combined with > 1 SD above the mean on errors were 
coded as “0” and excluded from further analysis for that particular task.  This included 4 
males and 3 females on the ANT, 1 male on the Counting Stroop task, and 2 males and 1 
female on the Emotion task. 
Attention task RTs were related across tasks, as were errors (Tables 27-29).  In 
addition, Principal Component factor analysis with Varimax rotation of RT and error 
scores revealed two factors that together explained 57.89% of the variance in scores.  The 
first consisted of RT data across tasks, while the other consisted of error data across 
tasks.  These findings suggested that task performance was related in a consistent fashion, 
and a composite score across tasks (Total Interference) was calculated as the mean of the 
individual task scores.  Thus, an individual whose score on the ANT was zero would 
receive a composite score that only included the other tasks. 
 Means and standard deviations of Total Interference scores by gender and risk 
group are presented in Table 30.  A 2 x 3 (Gender x Risk Group) univariate ANOVA 
revealed a significant main effect of risk group (F(1, 91) = 3.66, p < .05), such that 
individuals in the higher risk groups had higher interference scores.  No gender 
differences were found.  Table 31 presents correlations between the adjusted interference 
scores.  Counting Stroop Interference and ANT Interference were positively correlated.  
131 
 
In addition, Counting Stroop and Emotion Stroop were correlated at trend level.  
However, Emotion Stroop and ANT Interference were not related. 
 
 
TABLE 30.  Total Interference Scores by Risk Group and Gender 
 
Risk Group Gender N Mean (SD) 
Male 17 2.10 (.66) 
No Risk 
Female 10 2.15 (.65) 
Male 15 2.67 (.71) 
At Risk 
Female 17 2.12 (.52) 
Male 12 2.58 (.53) 
High Risk 
Female 28 2.64 (.90) 
 
 
 
 
TABLE 31.  Pearson’s Correlations between 
Adjusted Interference Scores 
Genders Combined 
 
 
  1 2 3 
 1-ANT -  
 
 
 2-CT Stroop .21* - 
 
 3-Emotion .05 .16 -
 
 
 
132 
 
 
 
Correlations between Measures 
 
 
Temperament and Attention Tasks 
 
 
 Correlations between the adjusted attention interference scores and both self- and 
mother-report temperament with genders combined are reported in Tables 32-33.  Results 
are summarized here by scale. 
 
Effortful Control 
 
 
There were no significant relationships between the adjusted interference scores 
and self-report temperament.  However, adjusted ANT interference was negatively 
related to mother-report Effortful Control including Inhibitory Control and, at trend level, 
Attention.  Stroop Interference was also negatively related to mother-report Effortful  
Control, including Activation Control and, at trend level, Inhibitory Control.  Adjusted 
Emotion Stroop Interference was also negatively related to mother report Inhibitory 
Control.  The Total Interference measure was negatively related to mother-report 
Effortful Control, including Attention and Inhibitory Control. 
 
Affiliativeness 
 
 
No significant relations between interference and Affiliativeness and its subscales were 
observed. 
133 
 
Negative Affectivity 
 
 
 ANT interference was negatively related to self-reported Frustration, Depressive 
Mood at trend level, and overall Negative Affectivity. However, mother-report ANT and 
Counting Stroop Interference were positively related to Aggression, as was Total 
Interference.  Counting Stroop and Emotion Interference were related, at trend level, to 
Depressive Mood, and Total Interference was also related to Depressive Mood.  Counting 
Stroop, Emotion, and Total Interference were all positively related to overall mother-
report Negative Affectivity. 
 
Surgency 
 
 
No significant relations existed between self-report or mother-report Surgency 
and performance on the attention tasks. 
 
Adjusted Interference Scores and Father-report Temperament 
 
 
 A subsample of 40 participants also had father-report temperament data, allowing 
comparison of attention task performance and father-report temperament scores.  Results 
are presented in Table 34.  ANT interference was negatively related to father-report 
Inhibitory Control and, at trend level, to both Attention and overall Effortful Control.  In 
addition, ANT interference was positively related to father-report Depressive Mood and, 
at the level of a trend, overall Negative Affectivity.  Total Interference was related to 
father-report Depressive Mood at trend level.
134 
 
 
 
 
 
TABLE 32.  Pearson’s Correlations – Adjusted Interference Scores  
and Self-Report Temperament, Genders Combined 
 
 
 ANT Stroop Emotion Total Interference Interference Interference Interference 
Effortful Control .12 .00 .09 .10 
Activation Control .11 .00 .14 .11 
Attention .10 -.08 -.00 .01 
Inhibitory Control .05 .07 .06 .10 
Affiliativeness -.06 -.06 -.00 -.08 
Affiliation -.07 -.12 -.03 -.14 
Perceptual Sensitivity -.12 -.04 .02 -.06 
Pleasure Sensitivity .05 -.01 .01 .01 
Negative Affectivity -.24* -.06 -.02 -.14 
Aggression -.01 -.07 .04 .02 
Depressive Mood -.17+ -.10 -.00 -.13 
Frustration -.21* .10 -.04 -.08 
Surgency .14 .09 -.05 .12 
Fear -.16 -.14 .03 -.15 
High Intensity Pleasure .07 .17 .04 .14 
Shyness -.09 .06 .10 .01 
+ p < .10, * p < .05, ** p < .01. 
 
135 
 
 
 
TABLE 33.  Pearson’s Correlations - Adjusted Interference Scores 
and Mother-Report Temperament, Genders Combined  
 
 
 ANT Stroop Emotion Total Interference Interference Interference Interference 
Effortful Control -.24* -.23* -.14 -.31** 
Activation Control -.09 -.22* -.02 -.16 
Attention -.20+ -.15 -.09 -.24* 
Inhibitory Control -.29** -.18+ -.23* -.36** 
Affiliation -.04 .09 -.08 .01 
Negative Affectivity .14 .21+ .22* .30** 
Aggression .21+ -.23* .14 .30** 
Depressive Mood .13 .19+ .19+ .26* 
Frustration -.04 .03 .14 .08 
Surgency -.15 .02 .11 -.01 
Fear .06 -.03 .04 .04 
High Intensity Pleasure -.16 .13 .15 .07 
Shyness .10 .09 -.12 .03 
+ p < .10, * p < .05, ** p < .01 
 
 
136 
 
 
TABLE 34.  Pearson’s Correlations - Adjusted Interference Scores 
and Father-Report Temperament, Genders Combined 
 
 
 ANT Stroop Emotion Total 
Interference Interference Interference Interference 
Effortful Control -.30+ .08 -.05 -.11 
Activation Control -.11 .11 .04 .03 
Attention -.30+ -.01 .01 -.13 
Inhibitory Control -.39* .10 -.18 -.21 
Affiliation -.20 -.15 .05 .03 
Negative Affectivity .28+ -.01 .13 .17 
Aggression .19 -.05 .16 .12 
Depressive Mood .42** .14 .08 .29+ 
Frustration .16 -.10 .09 .06 
Surgency -.12 -.11 .24 .02 
Fear .01 .05 -.26 -.11 
High Intensity Pleasure .01 .13 .25 .19 
Shyness .19 .24 -.02 .20 
 
+ p < .10, * p < .05, ** p < .01 
 
 
 
137 
 
 
 
TABLE 35.  Pearson’s Correlations – CFCQC and Temperament 
Genders Combined 
 
 
  Parent Conflict Risk Problem Prosocial Monitor Group Behavior Behavior 
Effortful Control .34** -.36** -.05 -.42** .39** 
Affiliativeness .23* .13 -.03 -.06 .30** 
Surgency -.07 .03 -.06 .16 .13 
Depressive 
Mood -.17 .21* .13 .15 -.10 
Aggression -.36** .28** .23* .36** -.31** 
Frustration -.07 .31** .10 .15 -.00 
M-Effortful 
Control .26* -.33** -.14 -.38** .21* 
M-Affiliation .14 -.09 .09 .01 .18 
M-Surgency .06 .09 .05 .04 .07 
M-Depressive 
Mood -.15 .18+ .01 .17 -.06 
M-Aggression -.17 .25* .05 .16 -.11 
M- Frustration -.07 .16 -.00 .07 -.03 
M = parent-report  
* p < .05, ** p < .01 
 
 
 
 
138 
 
 
 
 
 
TABLE 36.  Pearson's Correlations - Adjusted Interference Scores, 
CFCQC, and Risk Group 
Genders Combined 
 
 
Scale ANT Counting Emotion Total Stroop Stroop Interference 
Parental Monitoring -.06 -.12 -.09 -.16 
Parental Rules .04 .16 -.05 .05 
Family Conflict .11 .14 .10 .15 
Relationship Quality .03 -.16 -.08 -.09 
Deviant Peer Associations .21* .13 .21* .28** 
Deviant Behaviors .27** .12 .10 .24* 
Pro-social Peer Associations .11 -.14 -.04 -.05 
Pro-social Behaviors .04 -.10 .04 .02 
Risk Group .15 .19+ .19+ .30** 
+ p < .10;* p < .05; ** p < .01 
 
 
 
 
 
 
 
139 
 
 
TABLE 37.  Multiple Regression Analyses Dependent Variable: 
Problem Behavior Composite 
 
 
Model R2 1 Sig. 
   Beta t Sig. 
 .01 .464   Risk Group .08 .18 .464
Model 
2a R
2 Sig. R2∆ Sig.  Beta t Sig. 
 .24 .000 .23 .000 Family Conflict .31 2.88 .005
     Parental Monitoring -.33 -2.67 .009
     Risk Group .02 .17 .861
     Prosocial Behavior .10 .85 .396
Model 
2b R
2 Sig. R2∆ Sig.  Beta t Sig. 
 .38 .000 .37 .000 Effortful Control -.46 2.78 .007
     Mother Effortful Control -.25 -2.72 .008
     Total Interference .25 2.78 .007
     Surgency .16 1.78 .079
Model 
3 R
2 Sig. R2∆(a) Sig.  Beta t Sig. 
 .45 .000 .21 .000 Family Conflict .17 1.73 .088
     Parental Monitoring -.23 -2.10 .039
     Risk Group -.10 -1.11 .270
     Prosocial Behavior .19 1.75 .084
     Effortful Control -.39 -4.09 .000
     Mother Effortful Control -.19 -2.03 .046
     Total Interference .27 2.95 .004
     Surgency .09 1.07 .289
     
 
140 
 
TABLE 38.  Multiple Regression Analyses 
Dependent Variable:  Depressive Mood 
 
Model 1 R2 Sig.    Beta t Sig. 
 .02 .172   Risk Group .14 1.38 .172
Model 2a R2 Sig. R2∆ Sig.  Beta t Sig. 
 .06 .384 .04 .362 Family Conflict .15 1.23 .223
     Parental -.01 -.07 .946
Monitoring 
     Risk Group -.137 1.25 .216
     Problem Behavior .050 .41 .680
     Prosocial -.03 -.20 .842
Model 2b R2 Sig. R2∆ Sig.  Beta t Sig. 
 .48 .000 .47 .000 Gender .26 2.69 .009
     Effortful Control -.34 -4.16 .000
     Affiliativeness .23 2.67 .009
     Total Interference -.03 -.40 .692
     Surgency -.35 -4.48 .000
Model 3 R2 Sig. R2∆(a) Sig.  Beta t Sig. 
 .52 .000 .46 .000 Family Conflict -.01 -.09 .931
     Parental -.09 -.91 .368
Monitoring 
     Risk Group .05 .559 .578
     Problem Behavior .11 1.12 .267
Prosocial 
     -.06 -.53 .598
Behavior 
     Gender .27 3.13 .002
     Effortful Control -.26 -2.72 .008
     Affiliativeness .25 2.86 .005
     Surgency -.38 -4.53 .000
     Total Interference -.09 -1.00 .598
141 
 
 
 
Summary of Correlations 
 
 
 In general, self-reported Effortful Control was not related to attention 
interference.  However, both mother- and father-reported Effortful Control showed a 
negative relationship with attention interference, indicating that poor performance on the 
tasks was associated with lower levels of parent-reported Effortful Control.  Self-reported 
Negative Affectivity was somewhat negatively related to attention interference.  
However, parent-report Negative Affectivity was positively related to attention 
interference, indicating that poor performance on the tasks was associated with higher 
levels of negative affect.  There was little relationship between Affiliativeness and 
attention and Surgency and Attention. 
 
Temperament and Child and Family Center Questionnaire - Child Form (CFCQC) 
 
 
 Correlations between CFCQC scores and self- and mother-report temperament 
scores are reported in Table 35, and are summarized here by temperament scale. 
 
Effortful Control 
 
 
Parental Monitoring, and Relationship Quality were positively related to both 
self- and mother-report Effortful Control, while Parental Rules was also related to self-
report Effortful Control.  Family Conflict was negatively related to mother-report 
Effortful Control, and Deviant Peers and Deviant Behaviors were both negatively related 
to both self- and mother-report Effortful Control.  Prosocial Peers and Prosocial Behavior 
142 
 
scores were positively related to self- report Effortful Control and, at trend level, to 
mother-report Effortful Control. 
Affiliativeness 
 
 
 Parental Monitoring was positively related to self-report Affiliativeness and, at 
trend level, to mother-report Affiliativeness.  No other significant relations were found 
between Affiliativeness and CFCQC scores. 
 
Negative Affectivity 
 
 
 Parental Monitoring, Parental Rules, and Relationship Quality were all negatively 
related to self-reported Aggression, while Parental Monitoring and Relationship Quality 
were also negatively related to mother-report Aggression.  Deviant Behavior and Deviant 
Peer scores were positively related to self-, but not mother-report of Aggression.  In 
addition, Prosocial Behavior scores were negatively correlated with both mother- and 
self-reported Aggression.  Risk Group was positively related to self-, but not mother-
report of Aggression. 
 Relationship Quality and Parental Rules were both negatively related to self-
report Depressive Mood.  Deviant Behavior scores were related to self-reported 
Frustration, while Relationship Quality was negatively related to self-reported 
Frustration. 
143 
 
TABLE 39.  Multiple Regression Analyses 
Dependent Variable:  Aggression 
 
Model R2 1 Sig. 
   Beta t Sig. 
 .08 .009   Risk Group .27 8.70 .009
Model 
R2 Sig. R2∆ Sig.  Beta t Sig. 
2b 
 .27 .000 .19 .000 Family Conflict .07 .63 .530
     Parental Monitoring -.05 -.43 .665
     Risk Group .27 2.77 .007
     Prosocial Behavior -.23 -2.02 .047
     Problem Behavior .23 2.17 .033
Model 2 2  
1b R Sig. R ∆ Sig.  Beta t Sig. 
 .26 .000 .18 .000 Effortful Control -.43 -4.59 .000
Mother Effortful 
     -.14 -2.41 .018
Control 
     Total Interference -.05 -.51 .610
Model R2 Sig. R2∆(a) 2 Sig.  Beta t Sig. 
 .36 .015 .09 .000 Family Conflict .02 .18 .858
     Parental Monitoring -.04 -.29 .775
     Risk Group .27 2.85 .005
     Pro-Social Behavior -.15 -1.26 .210
     Problem Behavior .17 1.43 .155
     Self Effortful Control -.26 -2.30 .024
Mother Effortful 
     -.14 -2.30 .185
Control 
     Total Interference .17 -1.65 .102
144 
 
 
 
TABLE 40.  Multiple Regression Analyses 
Dependent Variable:  Pro-Social Behavior Composite 
 
Model R2 Sig.  1   Beta t Sig. 
 .01 .450   Risk Group .08 .76 .450
Model 
2a R
2 Sig. R2∆ Sig.  Beta t Sig. 
 .38 .000 .37 .000 Family Conflict -.17 -1.70 .092
     Parental Monitoring .56 5.54 .000
     Risk Group .16 1.83 .070
     Problem Behavior .08 .85 .396
Model 2 2  
2b R  Sig. R ∆ Sig.  Beta t Sig. 
 .23 .000 .21 .000 Effortful Control .45 4.62 .000
     Mother Affiliation .24 2.52 .014
     Total Interference -.06 -.61 .543
Model 
3 R
2 Sig. R2∆(a) Sig.  Beta t Sig. 
 .45 .000 .07 .000 Family Conflict -.12 -1.25 .215
     Parental Monitoring -.23 -2.10 .039
     Risk Group .17 1.99 .050
     Problem Behavior .19 1.75 .084
     Effortful Control .31 3.06 .003
     Mother Affiliation .13 1.54 .128
     Total Interference -.09 -1.01 .316
 
145 
 
 
 
TABLE 41.  Pearson’s Correlations – CFCQC and Temperament 
Females 
 
 
  Parent Conflict Risk Problem Prosocial Monitor Group Behavior Behavior 
Effortful Control .30* -.31* -.09 -.35** .26+ 
Affiliativeness .23+ .10 -.12 .03 .33+ 
Surgency -.19 .21 -.00 .29* .09 
Depressive 
Mood -.12 .18 .11 .16 -.04 
Aggression -.32* .25+ .24+ .36** -.22 
Frustration -.04 .19 .13 .07 .04 
M-Effortful 
Control .18 -.34* -.00 -.35** .20 
M-Affiliation .13 .05 .08 .02 .06 
M-Surgency -.12 .23 .15 .05 -.03 
M-Depressive 
Mood -.10 .16 -.01 .20 -.06 
M-Aggression -.13 .27* -.06 .04 -.11 
M- Frustration .07 .09 -.19 -.10 .09 
M = Mother-report 
* p < .05, ** p < .01 
146 
 
 
 
TABLE 42.  Pearson’s Correlations – CFCQC and Temperament 
Male 
 
 
  Parent Risk Problem Prosocial Monitor Conflict Group Behavior Behavior 
Effortful Control 39* -.43** -.04 -.51** .55** 
Affiliativeness .23 .19 -.12 -.11 .18 
Surgency .15 -.29+ -.04 -.12 .27+ 
Depressive 
Mood -.35* .33* -.06 .30+ -.40* 
Aggression -.43** .33* .24 .36* -.44** 
Frustration -.14 .50** .01 .30+ -.10 
M-Effortful 
Control .35* -.33* -.37* -.40* .18 
M-Affiliation .15 -.34 .03 .04 .29+ 
M-Surgency .31+ -.12 .04 -.02 .27 
M-Depressive 
Mood -.30+ .25 -.11 .21 -.19 
M-Aggression -.23 .21 .19 .31+ -.10 
M- Frustration -.24 .27+ .29+ .26 -.12 
M = Mother-report 
* p < .05, ** p < .01 
147 
 
 
Surgency 
 
 
 Self-reported Surgency was positively related to Prosocial Behavior scores.  
Mother-reported Surgency was negatively related to Relationship Quality at trend level. 
 
General Summary of Correlations 
 
 
 In general, both self- and parent-report Effortful Control were positively related to 
positive parenting, relationship, and prosocial behavior scores, and negatively related to 
problems with deviant peers and behaviors.  Depression, Frustration, and Depressive 
mood were all negatively related to Relationship Quality.  Aggression and Depressive 
Mood were also negatively related to Prosocial Behaviors. 
 Self-reported Surgency was associated with self-reported Prosocial Behaviors, but 
mother-reported Surgency was related to Family Conflict.  Affiliativeness was associated 
with Parental Monitoring. 
 
Attention Tasks and Child and Family Center Questionnaire - Child Form 
 
 
 Relationships between adjusted interference scores and CFCQC scores are 
presented in Table 36.  Risk Group is also included.  Deviant Behavior scores were 
positively related to ANT Interference and the Total Interference score.  Deviant Peer 
Relation scores were positively related to ANT, Emotion, and Total Interference Scores.  
In addition, Risk Group was positively associated with Total Interference Scores and, at 
trend level, with Emotion and Counting Interference scores. 
 
148 
 
Multivariate Analyses 
 
 
 As Deviant Peer and Deviant Behavior scores were highly correlated, multivariate 
ANOVA was performed to examine the effect of CFCQC, temperament, and risk group 
variables on the combined set of dependent variables.  Three separate analyses were run; 
the first examined Risk Group alone, as middle school Risk Group had been predicted to 
be indicative of future deviant activities.  However, Risk Group was not related to the 
combined set of variables (F(4, 184) = 1.44, ns). 
 In the second analysis, the remaining CFCQC variables were used.  The 
multivariate test revealed that Parental Monitoring was significantly related to the 
combined set of deviant behavior variables (F(2, 88) = 4.19, p < .05), as were, at trend 
level, both Relationship Quality (F(2, 88) = 2.88, p = .06) and Parental Rules (F(2, 88) = 
2.90, p = .06).  Univariate tests revealed Parental Monitoring to be significantly related to 
Deviant Behaviors (F(1, 89) = 8.42, p < .05) and, at trend level, to Deviant Peer Relations 
(F(1, 89) = 3.58, p = .06).  In addition, Parental Rules was related to Deviant Peer 
Relations (F(1, 89) = 5.169, p < .05).  Relationship Quality was related to Deviant 
Behaviors (F(1, 89) = 5.75, p < .05) and, at trend level, to Deviant Peer Relations (F(1, 
89) = 3.58, p = .06).  Overall, R2 =.273 for Deviant Behaviors, and .284 for Deviant Peer 
Relations. 
 In the third analysis, temperament variables were used.  Multivariate tests 
revealed a significant relationship between self-reported Effortful Control and the 
combined set of dependent variables (F(2, 69) = 6.12, p < .01).  In addition, self-reported 
Affiliation was related to the combined set of variables at the level of a trend (F(2, 69) = 
149 
 
2.54, p = .09).  Examination of univariate results revealed that self-reported Effortful 
Control was related to both Deviant Behaviors (F(1, 70) = 11.97, p < .01) and to Deviant 
Peer Associations (F(1, 70) = 8.52, p < .01), while self-reported Affiliation was only 
related to Deviant Behaviors (F(1, 70) = 3.20, p = .08).  Overall, R2 = .446 for Deviant 
Behaviors, and .392 for Deviant Peer Relations. 
 
Regression Analyses to Predict Psychosocial Outcomes 
 
 
 The multivariate analyses indicated that, overall, the set of individual difference 
variables was more highly related to Deviant Behavior and Deviant Peer scores than was 
the set of CFCQC variables, as indicated by R2.  To understand better the relative 
contribution of CFCQC, risk, temperament, and attention variables to psychosocial 
outcomes when examined together, a series of regression analyses were performed.  
However, the ratio of variables to subjects was too high to include all variables studied.  
In addition, the inclusion of both Deviant Peer and Deviant Behavior scores as predictors 
of other variables would have created problems with collinearity.  Therefore, a series of 
steps were taken before regression analyses were performed. 
First, CFCQC variables were combined as follows: a mean of Deviant Peers and 
Deviant Behavior scores was calculated and combined to form a Problem Behavior 
Composite; a mean of Prosocial Peers and Prosocial Behavior Scores was calculated and 
combined to form a Prosocial Behavior Composite; a mean of Parental Monitoring and 
Parental Rules was calculated and named Parental Monitoring/Rules; and a mean of 
Family Conflict and reverse-scored Relationship Quality scores was calculated and 
labeled Family Conflict/Relationships. 
150 
 
Next, a series of regressions were conducted in which all temperament scores and 
gender were entered to predict  scores on 1) Problem Behavior; 2) Prosocial Behavior; 3) 
Aggression; and 4) Depressive Mood.  Only those variables predicting significant 
variance in the dependent variables were retained for further analysis.  For each of the 
four dependent variables, hierarchical regression was performed with Risk Group entered 
in one block, CFCQC composite variables entered as another set of variables, and Total 
Interference and appropriate temperament scores, along with gender (if applicable) as a 
third set of variables.  The variables within each group were entered simultaneously.  
Two regression analyses were conducted for each dependent variable; the first contained 
Risk Group in the first block, CFCQC variables in the second block (2a), and 
temperament, Total Interference scores and gender (if applicable) in the third block.  The 
second regression for each variable contained Risk Group in the first block,  
temperament, Total Interference, and gender (if applicable) in the second block (2b), and 
CFCQC variables in the third. 
 
Problem Behavior Composite 
 
 
 Results for the Problem Behavior Composite are presented in Table 37.  For the 
first regression, Risk group did not predict significant variance in Problem Behavior 
scores.  However, Block 2 variables (CFCQC) were significant (r2 = .24, r2 ∆ = .24, F ∆ 
(3, 85) = 5.66, p < .001).  Examination of standardized Beta weights revealed that both 
Parental Monitoring/Rules (t = -2.67, p < .01) and Family Conflict/Relationships (t = 
2.87, p < .01) made significant contributions to the prediction of Problem Behavior 
scores. 
151 
 
When Block 3 variables were entered, significant additional variance was gained 
(r2 ∆ = .21, F Change (4, 81) = 7.589, p < .001).  The overall model remained significant 
(r2 = .45, F(8, 81) = 8.29, p < .001).  While Parental Monitoring still contributed 
significant variance, Family Conflict dropped to trend level.  Both self-report Effortful 
Control (t = -4.09, p < .001) and mother-report Effortful Control (t = -2.02, p < .05) 
contributed significantly to prediction of Problem Behavior scores, as did Total 
Interference (t = 2.948, p < .01).  The Pro-social composite measure contributed variance 
as well (t = 1.75, p = .08), but only at trend level.  The direction of this relationship 
indicated that Pro-social and Problem Behavior scores were positively related when 
controlling for the additional variables in the model. 
 
TABLE 43.  Pearson’s Correlations – Adjusted Interference Scores, CFCQC, and Risk 
Genders Combined 
 
 
  Parent Risk Problem Prosocial Monitor Conflict Group Behavior Behavior 
ANT -.07 -.03 .15 .30** .13 
Stroop -.02 .16 .19+ .13 -.14 
Emotion -.12 .09 .19+ .20+ .01 
Total -.13 .10 .30** .30** -.01 
+ p < .10, * p < .05, ** p < .01 
 
152 
 
 
 
TABLE 44.  Pearson’s Correlations – Adjusted Interference Scores, 
CFCQC, and Risk 
Females 
 
 
  Parent Conflict Risk Problem Prosocial Monitor Group Behavior Behavior 
ANT -.01 .00 .08 .44** .18 
Stroop -.15 .16 .27* .26+ -.25+ 
Emotion -.14 .12 .27* .19 -.08 
Total -.14 .10 .31* .38** -.09 
   + p < .10, * p < .05, ** p < .01 
 
 
 
TABLE 45.  Pearson’s Correlations – Adjusted Interference Scores, 
CFCQC, and Risk 
Males 
 
 
  Parent Conflict Risk Problem Prosocial Monitor Group Behavior Behavior 
ANT -.16 -.08 .27 .08 .07 
Stroop .15 .18 .11 -.02 -.03 
Emotion -.09 .04 .12 .22 .14 
Total -.11 .09 .30+ .19 .11 
+ p < .10, * p < .05, ** p < .01 
 
 
153 
 
 Regression 2 for Problem Behavior was then conducted.  Block 2a variables 
(temperament and Total Interference) were significant (r2 = .38, F(4, 85) = 13.13, p < 
.001).  Self-report Effortful Control (t = -5.21, p < .001), mother-report Effortful Control 
(t = -2.72, p < .01) and Total Interference (t = 2.78, p < .01) all contributed significantly 
to prediction of Problem Behavior scores, as did Surgency (t = 1.78, p = .08), but only at 
the level of a trend.  After entry of the second block of variables, the model was identical 
to that in regression 1. 
 
Depressive Mood 
 
 
 Results for self-report Depressive Mood are presented in Table 47.  For the first 
regression, Risk Group did not significantly predict Depressive Mood.  The addition of 
Block 2 (CFCQC) variables did not add significantly to the prediction of Depressive 
Mood scores.  When Block 3 variables were entered, significant additional variance was 
gained (r2 ∆ = .46, F Change (7, 77) = 11.67, p < .001).  The overall model then became 
significant (r2 = .52, F(12, 77) = 7.71, p < .001).  Examination of standardized beta 
weights revealed self-report Effortful Control (t = -2.72, p < .05), self-report 
Affiliativeness (t = 2.86, p < .05), self-report Surgency (t = -4.53, p < .001), and Gender 
(being female) (t = 3.13, p < .01) all contributed significantly to prediction of Depressive 
Mood. 
 Regression 2 for Depressive Mood was then conducted.  Block 2b variables 
(temperament and Total Interference) were significant (r2 = .48, F(6, 82) = 13.40, p < 
.001).  Self-report Effortful Control (t = -4.16, p < .001), Affiliativeness (t = 2.67, p < 
.01), and Surgency (t = -4.48, p < .001), as well as Gender (t = 2.69, p < .001), 
154 
 
contributed significantly to prediction of Depressive Mood scores.  After entry of the 
second block of variables, the model was identical to that in regression 1a. 
 
Aggression 
 
 
 Results of the regression analyses to explain self-report Aggression scores are 
presented in Table 39.  For regression 1, Risk Group predicted significant variance in 
Aggressive Mood scores (r2 = .08, F(1, 88) = 7.17, p < .01).  Block 2 variables added 
significantly to the prediction of Aggression scores (r2 ∆ = .20, F ∆ (4, 84) = 5.67, p < 
.001)).  Examination of standardized Beta weights revealed that Prosocial Composite 
scores (t = -2.02, p < .05), and Behavior Problem scores (t = 2.17, p = .03) made a 
significant contribution to the prediction of self-reported Aggression scores, such that 
Problem Behavior scores were positively related to Aggression and Prosocial behavior 
was negatively related to Aggression scores. 
When Block 3 variables were entered, small but significant additional variance 
was gained (r2 ∆ = .09, F ∆ (3, 81) = 3.70, p < .05).  The overall model remained 
significant (r2 = .36, F(7, 82) = 5.68, p < .001).  While Risk Group still contributed 
significant variance, Prosocial and Problem Behavior did not.  Self-report Effortful 
Control (t = -2.29, p < .05) contributed significantly to prediction of Aggression scores. 
 Regression 2 for Aggression was then conducted.  Block 2b variables 
(temperament and Total Interference) were significant (r2 = .26, F(3, 86) = 10.09, p < 
.001).  Self-report Effortful Control (t = -4.59, p < .001), and mother-report Effortful 
Control (t = -2.41, p < .05) both contributed significantly to the prediction of Aggression 
155 
 
scores.  After entry of the second block of variables, the model was identical to that in 
regression 1. 
 
Prosocial Composite 
 
 
 Results of the regression analyses to explain self-report Prosocial Behavior scores 
are presented in Tables 40.  For regression 1, Risk Group did not predict significant 
variance in Prosocial Composite scores.  Block 2 variables were significant (r2 ∆ = .21, F 
∆ (3, 85) = 7.60, p < .001).  Examination of standardized Beta weights revealed that 
Parental Monitoring (t = 5.54, p < .001) made a significant contribution to the prediction 
of self-reported Prosocial behavior scores, as did both Family Conflict (t = -1.70, p = 
.09), and Risk Group (t = 1.84, p = .07), such that Family Conflict was negatively related 
to Prosocial behavior, while Risk Group was positively related. 
When Block 3 variables were entered, small but significant additional variance 
was gained (r2 ∆ = .07, F ∆ (3, 82) = 3.47, p < .05).  The overall model remained 
significant (r2 = .45, F(7, 82) = 9.57, p < .001).  All three significant predictors from 
Block 2 remained significant.  In addition, self-report Effortful Control was positively 
related to Prosocial Behavior scores (t = 3.06, p < .01). 
 Regression 2 for Prosocial Behavior was then conducted.  Block 2b variables 
(temperament and Total Interference) were significant (r2 = .23, F(3, 86) = 8.38, p < 
.001).  Self-report Effortful Control (t = 4.62, p < .001), and mother-report Affiliation (t = 
2.51 p < .05) both contributed significantly to explanation of Prosocial behavior scores.  
After entry of the second block of variables, the model was identical to that in regression 
1. 
156 
 
Parenting and Effortful Control 
 
 
 Figure 14 illustrates the additive nature of the relationship between parenting, 
Effortful Control, and problem behaviors.  Low Parental Monitoring represents the 
bottom 25% of scores on the variable, high Parental Monitoring represents the top 25%;  
 
 
 1.5
 
 1.0
 
 .5
 
 0.0
 Parental Monitoring
 -.5 -Low
 D Med
 -1.0 ---Lo-w-----Me-d ----- HighHig-h --. 
 
Mean Self/Mother Effortful Control
 
 
FIGURE 14.  Standardized Problem Behavior Scores  
by Levels of Effortful Control and Parental Monitoring 
 
Standardized Problem Behavior Scores
157 
 
the remainder of the scores are categorized as medium Parental Monitoring.  Self- and 
mother-report Effortful Control scores were combined, then categorized into 3 groups as 
well.  As illustrated, participants with both low levels of Parental Monitoring and low 
levels of Effortful Control were particularly likely to participate in problem behaviors.  
However, individuals high in Effortful Control with high Parental Monitoring showed the 
lowest overall levels of problem behaviors. 
 
Discussion 
 
 
 The primary goals of the current study were: 1) To examine the relationship 
between performance on tasks measuring executive attention and parent- and self-report 
Effortful Control; and 2) To examine the relative contribution of temperament, executive 
attention, and parenting, family, and risk variables to depressive mood and aggression, as 
well as to pro- and anti-social behaviors.  Secondarily, the current study allowed 
additional replication and extension of findings regarding the psychometric properties of 
the EATQ-R, specifically when used with a diverse sample of older adolescents.  Each of 
these goals will be discussed, followed by a summary discussion. 
 
EATQ-R 
 
 
Co-efficient alpha for scales and subscales in Study 2 were comparable, although 
slightly lower, to those found in Study 1.  Co-efficient alpha for the Inhibitory Control 
subscale was, again, low, indicating that the subscale may require additional revision. 
Agreement between reporters was also lower in Study 2 (.02 - .35, mean level of 
agreement: .21) than in Study 1 (.29 - .53, mean level of agreement: .40), and was 
158 
 
somewhat lower than expected from previous studies (Capaldi & Rothbart, 1992; Ellis & 
Rothbart, 2002).  Again, this may reflect a difference in age between the current sample 
and the younger samples used in previous studies, as Achenbach et al. (1987) report that 
agreement between reporters is significantly higher for 6- to 11-year-olds than for older 
adolescents. 
The level of agreement for Depressive Mood in Study 2 was higher than would 
have been expected from previous studies (Capaldi & Rothbart, 1992; Ellis & Rothbart, 
2002).  This level of convergence was almost entirely due to agreement between females 
and their mothers (.45 vs. .04 for males).  This suggests that mothers may be more aware 
of internalizing emotion in daughters than in sons, perhaps due to a greater female 
tendency to verbalize emotion, as has been suggested in studies of adolescent emotion 
(e.g., (Larson & Richards, 1994).  Father/self agreement for Depressive Mood was also 
relatively high in Sample 2 (.39); however, too few fathers responded to allow 
comparison of father/daughter versus father/son dyads.  In addition, mother/father 
agreement in Sample 2, while relatively good for both Depressive Mood and Aggression, 
was much higher for the former (.74) than for the latter (.46).  This finding is not in 
agreement with the Achenbach et al. (1987) analyses, who found better agreement 
between raters on measures of externalizing, versus internalizing behaviors. 
 Another interesting area of agreement and disagreement involved ratings of 
Frustration.  While mother/self Frustration ratings in Study 1 showed moderate 
agreement (.30), agreement in Study 2 was near zero (.02).  Further, there were no 
significant overall differences between ratings of self- and parent-report Frustration in 
males, and a statistically significant but fairly small difference between ratings of self- 
159 
 
and parent-report Frustration in females.  This suggests that direction of disagreement 
between self- and parent-report for Study 2 participants was not systematic; some 
participants reported themselves higher in Frustration than did their parents, while others 
reported themselves lower.  Agreement between parent-reporters was moderate (.40).  
The Frustration subscale showed good internal reliability in Study 2 for both self- (.82) 
and parent-report (.74).  Thus, there is no obvious explanation as to why convergence 
between self- and parent-report was so poor. 
 Agreement for Effortful Control subscales was also substantially higher for Study 
1 than for Study 2, particularly with regard to Activation and Inhibitory Control, although 
the poor convergence for the latter was primarily for females.  While Study 1 participants 
consistently reported themselves as higher in Effortful Control and its subscales than did 
their parents, again there was no systematic pattern of differences between parent- and 
self-report in Study 2 with regard to Activation and Inhibitory Control.  However, 
agreement between parent-reporters in Study 2 was quite high for Effortful Control (.75).  
Again, there is no obvious explanation as to the reason for the poor convergence. 
 Agreement between raters tended to be somewhat higher for female participants 
in this sample than for males.  Additionally, preliminary examination of agreement 
between raters by ethnicity indicated that agreement was higher for European-American 
participants than for African-American participants.  However, proportionally more of 
the at-risk and high-risk participants assessed thus far are African-American; thus, this 
finding may be confounded with risk group.  Further, sample sizes by ethnicity and risk 
group were too small to allow adequate analyses to determine if this was the case.  
Therefore, the results of this line of inquiry are not reported. 
160 
 
While level of agreement between reporters differed from Study 1 to Study 2, the 
relationship between scale scores was, for the most part, similar for both.  Parent-report 
Surgency and Affiliation were positively related in Study 2, while self-report Surgency 
and Affiliation were not, just as in Study 1.  However, self-report Affiliation and 
Depressive Mood were positively related in Study 2 participants, replicating a finding 
from our original EATQ-R study (Ellis & Rothbart, 2002).  Possible reasons for this 
relationship were discussed in Study 1; however, Affiliation and Depressive Mood were 
not related in parent-report (.04), again suggesting that the construct of Affiliation, as 
operationalized in our measure, is quite different for adolescents than for their parents.  
Further, Affiliation and Frustration were positively related in Study 2, but not in Study 1.  
This relationship may be due to the positive relationship of both to Depressive Mood.  
The pattern of relationships for the remaining self-report scales was quite similar in both 
studies; only the strength of the relationships differed. 
In sum, internal reliabilities for the EATQ-R in Studies 1 and 2 were similar to 
those found in our original EATQ-R study, with the exception of Inhibitory Control.  
Further, the relationships between scales were similar across the three studies, suggesting 
a relative degree of consistency in patterns of relationships amongst temperament 
variables. 
 
Attention Tasks 
 
 
 Analysis of ANT performance in Study 2 showed results very similar to those 
seen in adults (Fan et al., 2002).  Incongruent flankers resulted in significantly longer 
RTs than did congruent flankers across warning cue types (mean RT interference = 114 
161 
 
ms) as seen in adults (mean RT interference = 130 ms) and in children (mean RT 
interference = 153 ms), while Alerting scores (mean alerting = 22 ms) were similar to 
adult subjects (mean alerting = 34 ms) and less than that seen in 10-year-olds (mean 
alerting = 76 ms).  Orienting scores (mean orienting = 41 ms) were similar to those seen 
in adults (mean orienting = 66 ms) and 10-year-olds (mean orienting = 64 ms).  Further, 
while overall reaction times (mean RT = 629 ms) were somewhat longer than those seen 
in adults (mean RT = 550 ms), and somewhat shorter (mean RT = 699 ms) than those 
seen in 10-year-olds (Rueda & Posner, 2002???).  Additionally, the Interference, 
Alerting, and Orienting effects were not related within subjects with genders combined 
(see Table ??); however, Alerting and Orienting were related in females.  This has not 
been observed in other samples, and needs replication before we can speculate that a true 
gender difference exists. 
Another gender difference in the present study that has not been observed in other 
samples was the significantly lower RTs for males than for females.  However, after 
removal of five subjects who demonstrated unusually quick reaction times in 
combination with an unusually high number of errors, the difference was no longer 
significant, suggesting that the difference was an artifact of rapid, but inaccurate, 
performance on the part of a handful of male participants. 
These results suggest that, even though the task was performed outside of a 
laboratory setting in varying environments, the effects of the ANT are evident even in 
less than ideal testing conditions.  Tasks such as the ANT may prove to be appropriate for 
wide scale administration and need not be confined to small samples in the laboratory.  
162 
 
This could potentially prove beneficial in screening for difficulties in the executive 
attention system. 
Analyses of Counting and Emotion Stroop task data were also, for the most part, 
as expected.  Number trials reliably produced longer RTs than did neutral trials, 
indicating a conflict effect.  However, Emotion Interference tended to be normally 
distributed around zero, indicating that many participants actually had faster RTs for 
emotional than for neutral trials.  This unexpected finding was intriguing, suggesting that 
emotional stimuli may act as an alerting signal for many individuals.  Indeed, Emotion 
Interference scores were negatively related to ANT Alerting scores, primarily for 
females, suggesting that those individuals who most benefited from the alerting cue in the 
ANT task were also quicker to respond to emotion stimuli. 
The meaning of this finding is unclear; however, it may suggest that, for many 
females, emotional stimuli are very salient but are not necessarily troubling, causing a 
“jolt” to attentional systems.  However, for others, emotional stimuli may serve to capture 
attention, resulting in the longer reaction times seen in individuals with anxiety disorders.  
Interestingly, the Alerting Effect was positively related to self-report Effortful Control in 
females; however, it was not related to parent-report Effortful Control nor was there any 
relationship in males. 
Recoding of interference scores to include both RTs and errors revealed a 
relationship between ANT and Counting Stroop Interference.  This is theoretically 
important, as the two tasks have been shown to activate similar regions of ACC (Bush, 
Luu, & Posner, 2000).  Emotion Interference, on the other hand, was not related to ANT 
Interference and showed only a marginally significant relationship to Counting Stroop 
163 
 
Interference.  Again, this is of theoretical interest, in that Emotional Stroop tasks have 
been shown to activate a slightly different region of ACC than have more “cognitive” 
Stroop tasks.  However, examination of correlations by gender revealed that, for males, 
scores for the three interference tasks were not related and ANT and Emotion 
Interference scores were negatively related.  However, for females, ANT and Counting 
Stroop Interference were related, as were Counting and Emotion Interference.  These 
gender differences are intriguing and suggest the importance of further investigation.  
However, replication of such a finding is important before a true gender difference can be 
assumed. 
 
Temperament and Attention 
 
 
 With genders combined, adjusted ANT Interference was negatively related to 
self-report Depressive Mood and Frustration.  In addition, ANT and Counting Stroop 
Interference were positively related to mother-report Aggression, and negatively related 
to mother-report Effortful Control.  Counting and Emotion Interference were both 
positively related to mother-report Depressive Mood.  In addition, the Total Interference 
score was positively related to mother-report Aggression and Depressive Mood, and 
negatively related to mother-report Effortful Control. 
While the relationship between mother-report Effortful Control and adolescent 
task performance is in keeping with our work with toddlers, there was not a significant 
agreement between self-report Effortful Control and task performance.  This may indicate 
that mothers are better reporters of self-regulation than are adolescents themselves, at 
least in the current sample.  Mother-report of Negative Affectivity was also positively 
164 
 
related to interference scores, again, in keeping with our hypothesis, whereas self-report 
Negative Affectivity in males was negatively related to attention interference.  Again, this 
may suggest that mothers are better reporters of dimensions of temperament at this age 
than are adolescents. 
In sum, these findings provide an interesting link between functioning of the 
executive attention system and mother-report measures of self-regulation and emotion.  
We have found a similar relationship between mother-report temperament and toddler 
performance on executive attention tasks in two separate studies in our laboratory.  To 
find similar results across such samples that differ so greatly in age suggests that results 
of these tasks may, indeed, serve the role of markers of functioning of neural areas 
important in self-regulation. 
In addition, to find agreement between measures that vary substantially in 
methodology not only speaks to the validity of both measures, but supports the use of 
parent-report measures of temperament as relatively objective accounts of individual 
differences in children and adolescents.  Both mother- and father-report temperament 
related not only to each other, but to task performance in the hypothesized direction.  
This suggests that both mothers’ and fathers’ accounts of their children's behaviors and 
temperaments may provide valuable information regarding biologically based individual 
differences. 
 
Attention, Parenting, Risk, and Problem Behavior Variables 
 
 
 Adjusted interference scores were also related to risk group and CFCQC Deviant 
Behavior and Deviant Peer scores.  Specifically, ANT and Emotion Interference, as well 
165 
 
as Total Interference scores were both related to Deviant Peer Scores, and ANT and Total 
Interference scores were related to Deviant Behavior.  Counting and Emotion 
Interference were related to risk group, such that individuals in higher risk groups had 
higher Counting and Emotion Interference scores.  Indeed, univariate ANOVA revealed a 
significant main effect of risk group for Total Interference scores. 
These relationships suggest the importance of measuring attention regulation in 
adolescents.  That scores predict self-report association with deviant peers and 
involvement in deviant behaviors and distinguish between groups seen as “At Risk” by 
teacher ratings made three years previously gives some evidence of the validity of the 
approach.  However, poor task performance, Problem Behavior, and teacher ratings of 
risk could all reflect a general tendency for noncompliant behavior that is less reflective 
of actual functioning of neural systems and more reflective of differences in levels of 
rebellion. 
 
Temperament, Parenting, Risk, and Problem Behavior Variables 
 
 
Both self- and parent-report Effortful Control were positively related to self-report 
Parental Monitoring, Parental Rules, and Prosocial Behavior in this sample, and 
negatively to Family Conflict and Deviant Peer Relations and Deviant Behaviors.  Self-
reported Depressive Mood, Aggression, and Frustration were all related to Family 
Conflict, as were parent-report Depressive Mood and Aggression.  Further, self-report 
Aggression was related to risk group rating. 
That Effortful Control and parenting variables are positively related is not 
surprising.  Parenting is often found to covary with child characteristics (Putnam et al., in 
166 
 
press), thus suggesting that it may be easier to monitor children with good Effortful 
Control.  However, involved and conscientious parenting may also produce children with 
good self-regulation.  An additional explanation, however, is that the influence between 
parents and children is bi-directional, with parent characteristics influencing children's 
behavior and children's behavior influencing the behavior of parents.  This interpretation 
is supported by empirical research (e.g., Bates, Pettit, Dodge, & Ridge, 1998; Kochanska, 
1997; Patterson, DeBaryshe, & Ramsey, 1989; van den Boom, 1994). Early 
temperamental characteristics may “set in motion a chain of reactions from others that 
put children at risk or protect them from developing behavior and psychological 
problems” (Collins, Maccoby, Steinberg, Hetherington, & Bornstein, 2000, p. 222).  The 
current study does not allow examination of this process; such examination is only 
possible via careful, longitudinal research. 
Of additional interest is the prospect that passive gene-environment correlation 
may account for the relationship between parenting variables and child effortful control.  
A passive gene-environment correlation exists when similar behaviors between parents 
and children exist because both are genetically related (Deater-Decker, 2000).  We may 
posit that consistent, planful parenting that includes monitoring, rule making, and a 
positive relationship, may be due to high levels of effortful control on the part of the 
parent.  In this sense, then, we can think of parenting variables as a proxy for parent 
effortful control.  However, as parent effortful control was not measured in the present 
study, it is not possible to test this hypothesis directly. 
167 
 
Temperament, Attention, Parenting Variables and Psychosocial Outcomes 
 
 
 Regression analyses to examine variance in scores on psychosocial behaviors 
yielded results consistent with those hypothesized.  Parental Monitoring/Rules, self- and 
mother-report Effortful Control, and Total Interference scores all contributed 
significantly to the explanation of variance in Problem Behavior scores.  Further, there 
were no significant interaction effects, indicating that effects of each were independent 
and additive.  Within each level of Effortful Control, for example, Parental Monitoring 
had an effect.  Further, within each level of Parental Monitoring, Effortful Control had an 
effect, and so on.  Thus, participants with poor Effortful Control, poor executive attention 
skills, and poor Parental Monitoring had particularly high problem behavior scores. 
 However, it appears that parenting and family variables were partially mediated 
through their relationship with individual difference measures.  Parental 
Monitoring/Rules and Family Conflict were both significantly related to Problem 
Behavior scores when entered in the first block.  However, when individual difference 
measures were entered, the role of Family Conflict was somewhat diminished as was, to a 
lesser degree, the association with Parental Monitoring/Rules.  When individual 
difference measures were entered first, the effects of self- and mother-report Effortful 
Control were diminished slightly when family and risk variables were entered, but the 
contribution of Total Interference actually increased.  While both individual difference 
measures alone (r2 = .38) and family and risk variables alone (r2 = .24) predicted 
significant variance in problem behavior scores, inclusion of both sets of variables 
168 
 
provided a much more complete understanding of the processes involved in such 
behavior than did either set alone. 
Interestingly, family and risk variables did not predict significant variance in 
Depressive Mood scores above and beyond that predicted by individual difference 
measures alone.  This finding suggests that Depressive Mood may be more closely 
related to internal states than to environmental factors.  This interpretation would be 
consistent with a more biologically based interpretation of depression, rather than an 
interpretation which views depression as a reactive process triggered by environmental 
factors.  However, relationship issues have repeatedly been associated with depression 
(Frank & Young, 2002).  While relationship issues were not measured in the current 
study, the relation of Affiliativeness to Depressive Mood may reflect the importance of 
relationship issues. 
Total Interference scores did not contribute significantly to Depressive Mood 
scores, nor did Emotion Interference as hypothesized.  It is unclear why the hypothesized 
relationship was not found.  Further studies might include variants of emotional 
interference tasks to understand better any possible role of executive attention in 
depression. 
 Parental Monitoring/Rules and Family Conflict did not predict significant 
variance in Aggression scores, as would have been suggested by Patterson et al.'s theory 
(1992).  However, Prosocial scores were negatively related to Aggression, and Problem 
Behavior scores, along with risk group, were positively related to Aggression.  It was 
rather surprising that family variables did not contribute directly to the prediction of 
Aggression scores; however, partial correlations revealed that the relationships between 
169 
 
family variables and Aggression were mediated, in part, through their relationships with 
Prosocial and Problem Behaviors. 
 Interestingly, individual difference variables contributed less to the prediction of 
Prosocial behaviors than did family and risk group variables.  Surprisingly, while 
Parental Monitoring/Rules and Family Conflict were related to Prosocial Behavior scores 
in a direction that one would predict, both risk group and (after entering individual 
difference variables) Problem Behavior were positively related to Prosocial Behavior 
scores.  This relationship indicates that, after controlling for individual differences in 
Effortful Control, the same individuals engaging in problem behaviors are also engaging 
in prosocial behaviors.  In addition, this pattern is related to risk group.  This may be due 
in part to the type of questions asked in the prosocial composite, many of which involve 
participation in group activities.  Perhaps adolescents are more likely to both behave and 
misbehave when in a group; indeed, prosocial preschoolers also tend to exhibit 
aggressive behaviors. 
Fabes et al. (1999) in a meta-analysis of data focusing on the development of 
prosocial behavior, found such behaviors to increase during adolescence.  However, very 
little work has been done with respect to temperament and prosocial behaviors in 
adolescence.  Further, Fabes and colleagues suggest that the concurrent examination of 
both prosocial and antisocial behaviors is necessary, as examination of one without the 
other “presents a skewed and limited description of the complexity of adolescents” 
(p.13).  As such, additional work in these areas may be valuable. 
In Sum, Study 2 provided partial support for our hypotheses.  Performance on 
attention tasks was positively related to Effortful Control and negatively related to 
170 
 
Negative Affectivity, but only for parent-report.  This suggests that the aspects of 
temperament related to executive attention may be better assessed by parent-report in this 
age group than by self-report.  This finding will need to be replicated in additional 
samples.  In addition, parenting variables, self-regulation, and attention all contributed 
significant variance to explanation of problem behaviors as had been hypothesized, 
although Surgency did not.  Further, risk group did not contribute significant variance. 
Attention scores did not contribute significantly to explanation of Depressive 
Mood or Aggression, although temperament variables did.  Further, the pattern of 
temperament variables predicting Depressive Mood and Aggression were similar to those 
found in Study 1, including the role of Affiliativeness in Depressive Mood.  Individual 
difference scores did not add significantly to the prediction of prosocial behavior scores.  
The reason for this finding is not immediately clear. 
The general pattern of findings suggests the importance of self-regulation, as 
assessed by both questionnaire and laboratory measures, in psychosocial outcomes, 
particularly with regard to deviant peer relationships and behaviors.  These findings are in 
agreement with work by Giancola and others reporting deficits in executive functioning 
in adolescents involved in delinquent behaviors.  Further, these findings suggest that the 
link between executive functioning and delinquent behavior may be via Effortful Control, 
as suggested by our laboratory findings in this study as well as our studies with toddlers 
(Gerardi-Caulton, 2002; Rothbart et al., in press).  However, self- and mother-report 
Effortful Control scores and Total Interference scores all three predicted significant 
variance in Problem Behavior scores, suggesting that each also measure unique aspects of 
self-regulation important in our understanding of adolescent behavior. 
171 
 
In addition, the findings suggest that effective parenting and self-regulation go 
hand-in-hand in predicting outcomes.  These findings are consistent with current theory 
(e.g., Collins et al., 2001) highlighting the importance of both genes and environment in 
child outcomes and, again, underscore the importance of measurement of both types of 
variables in developmental studies. 
172 
 
 
 
 
 
CHAPTER IV 
 
CONCLUSIONS 
 
Overview 
 Two studies were designed to investigate the role of individual differences in 
adolescent psychosocial development, including temperament, attention, and puberty, and 
to relate temperamental Effortful Control to functioning of the executive attention 
system.  Study 1 sought to: 1) Examine possible relationships between temperament and 
puberty in early adolescence; and 2) Examine possible contributions of temperament and 
puberty to psychosocial outcomes in early adolescence.  Study 2 sought to: 1) Measure 
and examine possible relations between executive attention, temperament, and family 
variables; and 2) Examine relative contributions of executive attention, temperament, and 
risk, parenting, and family variables to psychosocial outcomes.  Both studies allowed 
further exploration of the psychometric properties of the EATQ-R. 
Relations were found between temperament and puberty, with puberty associated 
with an increase in aspects of negative affectivity.  Further, it appears that puberty may 
be associated with decreases in self-regulation.  In connection with temperament and 
psychosocial outcomes, we replicated in Study 1 our previous finding (Ellis & Rothbart, 
173 
 
2002) relating poor effortful control to both aggression and depressive mood, as well as a 
positive relationship between affiliation and depressive mood.  In addition, we found 
relationships between performance on measures of executive attention and parent-, but 
not self-reported temperament in Study 2, such that relatively poor performance on the 
tasks was related to lower levels of mother-report Effortful Control and higher levels of 
mother-report Negative Affectivity.  Further, task performance was related to teacher-
assigned risk categories, such that individuals rated five years previously as "at risk" or 
"high risk" for the development of deviant behaviors were more likely to perform poorly 
on the tasks. 
Additionally, individual difference variables were shown to play a significant role 
in self-report problem behaviors.  Both self- and mother-report Effortful Control, as well 
as performance on the attention tasks, were related to problem behaviors.  In addition, 
Effortful Control was negatively related to Aggression and Depression, as found both in 
Study 1 and in our original EATQ-R study (Ellis & Rothbart, 2002).  However, parenting 
variables also contributed significantly to both Problem and Prosocial behavior scores, 
but not to Aggression and Depressive Mood scores. 
We also found evidence of adequate internal reliability of the EATQ-R self- and 
parent-report measures, with the exception of the Inhibitory Control subscale.  We found 
higher than expected agreement between parent- and self-report in early adolescence, but 
somewhat poorer convergence in later adolescence.  However, we are not able to 
determine at this time if the poor convergence is primarily within a subgroup of 
participants.  Overall, while the results of the studies were encouraging, they must be 
interpreted with caution, as outlined in the limitations discussed below. 
174 
 
Limitations, Future Directions, and Conclusions 
 
 
 The idea that individual differences in temperament play an important role in 
adolescent psychosocial development is not a new one, nor is there a paucity of research 
on the role of puberty in adolescent development.  Further, there is ample empirical 
evidence to suggest that executive processes play a role in antisocial behaviors.  The 
current set of studies attempted to examine contributions from a number of different 
sources of individuality, including temperament, executive processes, and puberty.  In 
addition, temperament and executive processes were examined within the context of 
parenting and risk factors. 
In general, the studies supported the original hypotheses.  Individual differences, 
as assessed by self- and parent-report temperament measures, attentional marker tasks, 
and self- and parent-report pubertal status, all contributed in varying degrees to prediction 
of psychosocial development.  Further, individual differences in temperament and 
attention were seen to contribute to psychosocial outcomes above and beyond parenting, 
risk, and family variables.  Additionally, the studies supported the idea that the 
questionnaire and attentional measures developed in our laboratory are psychometrically 
sound and appropriate for use with a wide range of adolescents. 
Other results that emerged from the current studies were rather unexpected, such 
as the decline in Effortful Control and Fear across puberty, and convergence of attention 
task scores and parent-report temperament scores in the absence of convergence with 
self-report scores.  However, both studies suffered from limitations that require caution in 
interpretation of results.  First, the large number of correlations conducted in the analyses 
175 
 
raise the possibility that some findings labeled as significant may be the result of Type I 
error.  It has been suggested that multiple correlations be tested at a significance level 
equaling .05/C where C equals the number of correlations coefficients tested.  For 
instance, if one is testing 50 coefficients, they should be tested at .05/50 = .001 level.  
However, given that relationships in the social sciences are often not particularly strong, 
such an adjustment then creates a high risk of a Type II error.  As such, it is common for 
researchers to use .05 regardless of the number of coefficients tested; however, 
approximately 1 in 20 may be the result of error.  It is important to note, however, that 
the strength of the relationship does not change regardless of any adjustments for 
stringency of significance levels. 
An additional limitation has to do with the design of the studies.  Study 1 was 
cross-sectional in nature and, while Effortful Control and Fear were seen to decrease 
across pubertal development, a longitudinal study would be necessary to allow 
measurement of possible decreases within individual subjects.  Further, a decrease in 
Effortful Control would suggest a decrease in executive functioning; however, as 
attention was not assessed in Study 1, it is impossible to explore this possibility. 
Second, the measurement of secondary sexual characteristics as a proxy for the 
internal mechanisms of pubertal development can allow us to say very little about 
hormone levels, brain development, or other internal physiological changes hypothesized 
to play a role in psychosocial development.  While correlations between Tanner breast 
stage and estradiol levels are in the .60 to .80 range (Graber et al., 1996), there is no one-
to-one correspondence between secondary sexual characteristics and hormone levels.  In 
addition, the age range studied included too little variance in male pubertal development 
176 
 
to fully examine possible changes in emotional reactivity.  Most males studied were 
either pre-pubertal or in the very early stages of pubertal development. 
While participants in Study 2 are part of a long-term longitudinal study, measures 
of temperament and attention were not obtained in early adolescence; thus, it is not 
possible to examine scores for possible changes across time.  Further, as participants in 
Study 2 are 16- and 17-years-of-age, it is safe to assume that most females are post-
menarcheal, thus making the study of pubertal status, in the females at least, a moot 
point.  Additionally, the poor convergence between self-report temperament and attention 
scores is troubling; nonetheless, self- and parent-report Effortful Control, as well as 
attention scores, were significant in explaining variance in problem behavior, suggesting 
that all three are measuring important aspects of self-regulation.  However, problem 
behavior was measured via self-report, perhaps causing inflated relationships between 
other self-report variables and the outcome measure due to shared method variance. 
Neither study assessed life events and, in particular, relationship events, in the 
study of Depressive Mood, yet stressful life events have been implicated in a number of 
different studies of adolescent depression (see Petersen et al., 1993, for a review).  Thus, 
while individual difference variables do seem to play a significant role in depression, 
examining individual differences within the context of known environmental risk factors 
would provide a much richer framework of investigation. 
Nonetheless, the current set of studies may serve as an important stepping-stone 
in a future line of work.  Ideally, a group of adolescent participants would be followed 
through puberty and into late adolescence, with repeated measurement of attention, 
pubertal, and temperament variables in conjunction with environmental variables.  This 
177 
 
design would allow examination of possible changes in temperament and attention 
scores, and would also allow examination of such changes within the context of 
environment.  In addition, parents would be asked to provide periodic reports of their 
adolescents’ temperament.  This would allow exploration of the possibility that it is not 
temperament per se that changes during puberty, but rather the reporting tendencies of 
adolescents.  
 In conclusion, the current set of studies provides new information in two areas.  
First, pubertal maturation may be associated with a decrease in functioning of 
physiologically based regulation systems.  This finding has important implications in the 
study of adolescent psychosocial development, as poor self-regulation has been 
implicated in a number of studies of adolescent deviance.  Second, pubertal maturation 
may be related to increases in sensation seeking as well as in depressive mood.  These 
findings suggest that such increases may be based, to some degree, in physiological 
changes rather than changes in social systems.  Third, attention regulation, as measured 
by neural marker tasks, provides an important additional level of investigation of 
variables important in adolescent psychosocial outcomes.  Indeed, such tasks allow us to 
make tentative links between brain function and real-world behaviors. 
 Human development involves social, cognitive, physiological, cultural, and 
environmental consistencies and inconsistencies.  The variables that influence 
psychosocial development come from each of these domains, and interact with one 
another across the lifespan in intricate processes that differ from individual to individual.  
We may never fully understand the pathways that lead to positive outcomes in one child 
and negative outcomes in another.  However, by utilizing an integrative approach to the 
178 
 
study of development, we may in the future come to have a better understanding of the 
complex organisms that are our children. 
179 
 
APPENDIX A 
 
 
EATQ-R 
 
 
Self-Report Items By Scale 
 
Activation Control 
I have a hard time finishing things on time.  (R) 
I do something fun for awhile before starting my homework, even when I’m not 
supposed to.  (R) 
If I have a hard assignment to do, I get started right away. 
I finish my homework before the due date. 
I put off working on projects until right before they're due.  (R) 
Affiliation 
I want to be able to share my private thoughts with someone else. 
I enjoy exchanging hugs with people I like. 
I will do most anything to help someone I care about. 
It is important to me to have close relationships with other people.   
I am quite a warm and friendly person.   
Aggression 
If I'm mad at somebody, I tend to say things that I know will hurt their feelings. 
When I am angry, I throw or break things. 
If I get really mad at someone, I might hit them. 
I tend to be rude to people I don't like. 
When I'm really mad at a friend, I tend to explode at them. 
180 
 
I pick on people for no real reason. 
 
Attention 
It is easy for me to really concentrate on homework problems. 
I find it hard to shift gears when I go from one class to another at school.  (R) 
When trying to study, I have difficulty tuning out background noise and 
concentrating.  (R) 
I am good at keeping track of several different things that are happening around me. 
I pay close attention when someone tells me how to do something. 
I tend to get in the middle of one thing, then go off and do something else.  (R) 
Depressive Mood 
I feel pretty happy most of the day.  (R) 
My friends seem to enjoy themselves more than I do. 
It often takes very little to make me feel like crying. 
I get sad more than other people realize. 
I get sad when a lot of things are going wrong. 
I feel sad even when I should be enjoying myself, like at Christmas or on a trip. 
 
Fear 
I get frightened riding with a person who likes to speed. 
I worry about my family when I'm not with them. 
I worry about getting into trouble. 
I am nervous of some of the kids at school who push people into lockers and throw 
your books around. 
I worry about my parent(s) dying or leaving me. 
181 
 
I feel scared when I enter a darkened room at home. 
Frustration 
It bothers me when I try to make a phone call and the line is busy. 
I get very upset if I want to do something and my parents won't let me. 
I get irritated when I have to stop doing something that I am enjoying. 
It really annoys me to wait in long lines. 
I get very frustrated when I make a mistake in my school work. 
It frustrates me if people interrupt me when I'm talking. 
I get upset if I'm not able to do a task really well. 
Inhibitory Control 
It's hard for me not to open presents before I’m supposed to. (R) 
When someone tells me to stop doing something, it is easy for me to stop. 
The more I try to stop myself from doing something I shouldn't, the more likely I am 
to do it.  (R) 
It’s easy for me to keep a secret. 
I can stick with my plans and goals. 
Pleasure Sensitivity 
I like to feel a warm breeze blowing on my face. 
I enjoy listening to the birds sing. 
I like to look at the pattern of clouds in the sky. 
I like to look at trees and walk amongst them. 
I like the crunching sound of autumn leaves. 
 
 
182 
 
 
 
Perceptual Sensitivity 
I notice even little changes taking place around me, like lights getting brighter in a 
room. 
I tend to notice little changes that other people do not notice. 
I am very aware of noises. 
I can tell if another person is angry by their expression. 
Shyness 
I feel shy with kids of the opposite sex. 
I feel shy about meeting new people. 
I am shy. 
I am not shy.  (R) 
High Intensity Pleasure 
I think it would be exciting to move to a new city. 
I wouldn't like living in a really big city, even if it was safe.  (R) 
Skiing fast down a steep slope sounds scary to me.  (R) 
I would not be afraid to try a risky sport, like deep-sea diving. 
I wouldn't be afraid to try something like mountain climbing. 
I enjoy going places where there are big crowds and lots of excitement. 
 
183 
 
 
Parent-Report Items by Scale 
 
Activation Control 
Has a hard time finishing things on time.  (R) 
If having a problem with someone, usually tries to deal with it right away. 
Usually does something fun for awhile before starting her/his homework, even though 
s/he is not supposed to. 
When asked to do something, does it right away, even if s/he doesn't want to. 
Usually finishes her/his homework before it’s due. 
Usually gets started right away on difficult assignments. 
Usually puts off working on a project until it is due.  (R) 
Affiliation 
Likes taking care of other people. 
Likes to be able to share his/her private thoughts with someone else. 
Would like to be able to spend time with a good friend every day. 
Enjoys exchanging hugs with people s/he likes. 
Wants to have close relationships with other people. 
Is quite a warm and friendly person. 
Aggression 
When angry at someone, says thing s/he knows will hurt that person's feelings. 
If very angry, might hit someone. 
Tends to be rude to people s/he doesn't like. 
Slams doors when angry. 
184 
 
Doesn't criticize others.  (R) 
Tends to try to blame mistakes on someone else. 
Makes fun of how other people look. 
Attention 
Finds it easy to really concentrate on a problem. 
When interrupted or distracted, forgets what s/he was about to say.  (R) 
Has a difficult time tuning out background noise and concentrating when trying to study.  
(R) 
Is good at keeping track of several different things that are happening around her/him. 
Is often in the middle of doing one thing and then goes off to do something else without 
finishing it.  (R) 
Pays close attention when someone tells her/him how to do something. 
Depressive Mood 
Often does not seem to enjoy things as much as his/her friends. 
Feels like crying over very little on some days. 
Is sad more often than other people realize. 
Is hardly ever sad, even when lots of things are going wrong. (R) 
Sometimes seems sad even when s/he should be enjoying her/himself like at Christmas, 
or on a trip. 
Fear 
Worries about getting into trouble. 
Worries about our family when s/he is not with us. 
Is afraid of the idea of me dying or leaving her/him. 
Doesn't enjoy playing softball or baseball because s/he is afraid of the ball. 
Feels scared when entering a darkened room at night. 
185 
 
Is nervous being home alone. 
Frustration 
Is annoyed by little things other kids do. 
Gets very irritated when someone criticizes her/him. 
Gets irritated when I will not take her/him someplace s/he wants to go. 
Gets irritated when s/he has to stop doing something s/he is enjoying. 
Hates it when people don't agree with him/her. 
Gets very frustrated when s/he makes a mistake in her/his school work. 
Inhibitory Control 
Has a hard time waiting his/her turn to speak when excited.  (R) 
Opens presents before s/he is supposed to.  (R) 
Is more likely to do something s/he shouldn't do the more s/he tries to stop her/himself.  
(R) 
Is able to stop him/herself from laughing at inappropriate times. 
Is usually able to stick with his/her plans and goals. 
Shyness 
Can generally think of something to say, even with strangers. 
Is shy. 
Is not shy.  (R) 
Likes meeting new people. 
Feels shy about meeting new people. 
High Intensity Pleasure 
Thinks traveling to Africa or India would be exciting and fun. 
186 
 
Would be frightened by the thought of skiing fast down a steep slope.  (R) 
Thinks it would be exciting to move to a new city. 
Wouldn't be afraid to try a risky sport like deep sea diving. 
Expresses a desire to travel to exotic places when s/he hears about them. 
Would like driving a racing car. 
Likes it when something exciting and different happens at school. 
Is energized by being in large crowds of people. 
Wouldn't want to go on the frightening rides at the fair.  (R) 
 
187 
 
APPENDIX B 
 
BODY CHANGES QUESTIONNAIRES 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
188 
 
APPENDIX C 
 
STUDY 1 INSTRUCTION LETTER 
 
 
 
 
 
 
 
 
 
 
 
 
189 
 
APPENDIX D 
 
INSTRUCTION SCREENS – ATTENTION NETWORKS TASK 
 
Screen 1:  
This is an activity about attention.  You will see five arrows in the middle of the screen.    
Your job is to figure out which way the CENTER arrow is pointing.  You choose your 
answer by pressing the correct mouse button on the keyboard. 
   
For example, press the RIGHT button if the center arrow look like this: 
 
-> -> -> -> -> 
 
Screen 2:   
Sometimes the center arrow will point in the opposite direction as the arrows around it. 
Remember, you need to show which way the CENTER arrow is pointing.   
 
For example, the arrows might look like this: 
 
-> ->   <-   -> -> 
 
You should press the LEFT button because the CENTER arrow is pointing left. 
 
Screen 3:   
Before the arrows appear, there will be a plus sign ("+") in the center of the screen. The 
arrows will appear either above or below the ("+").  
 
Keep your eyes on the ("+") throughout the activity. 
 
          
Please press the SPACE bar to continue 
 
Screen 4:   
The Project Alliance assistant will show you where to place your hands.  You will start 
this activity by practicing several times.  Please try to work as quickly and accurately as 
possible. 
 
Please press the SPACE bar to begin. 
 
Post practice screen:   
You have completed the practice activity. 
190 
 
Now the rest of the activity can be started.  There are two more sessions in this activity.  
Each session takes about 5 minutes.   
 
 
When you are ready, press the SPACE bar to start the first session. 
 
Post session 1 screen:  
You have finished session 1. 
 
Please take a rest. 
 
 
To start the next session, press the SPACE key. 
 
Post session 2 screen: 
You are done with this activity. 
 
Please tell the PA staff member. 
191 
 
APPENDIX E 
 
CHILD AND FAMILY CENTER QUESTIONNAIRE 
 
Items by Scale 
 
Prosocial Behaviors   
 
- In the past 3 months, how often did you . . . 
 
Complete assignments or homework on time 
Participate in sports or organized activities (such as drama, music, religious groups) 
Help with chores around the house 
Cooperate with your parent(s) 
Cooperate with your teachers 
Attend church or religious activities 
Act friendly and helpful to family members 
Follow through with a plan or goal 
Feel positive about going to school 
Help someone not in the family 
Volunteer in the community 
Support a friend 
 
 
Parental Monitoring 
 
- In the past 3 months, how often did at least one of your parents . . . 
 
Know what you were doing when you were away from home 
Know where you were after school 
Know about your plans for the coming day 
Have a pretty good idea about your interests and activities 
Compliment you for anything you did well 
Give you something extra for doing something well 
 
 
Parental Rule Making 
 
- Fill in the blank with "Didn't have a rule or an expectation", "Sort of expected", 
"Definitely expected", "Had a clear rule" 
 
My parent(s) ________ that I should do homework every day 
My parent(s) ________ that I should not smoke cigarettes or use chewing (or smokeless) 
tobacco 
192 
 
My parent(s) ________ that I should not use alcohol 
My parent(s) ________ that I should not use marijuana 
My parent(s) ________ that I should not use drugs 
My parent(s) ________ that I should not be with friends at our house or someone else's 
house without an adult around 
 
Relationship Quality 
 
-Over the last month: 
 
I really enjoyed being with my parent(s) 
I got along very well with my parent(s) 
My parent(s) trusted my judgment 
There was a feeling of togetherness in our family 
Family members backed each other up 
Things our family did were fun and interesting 
Ate a meal with at least one of my parents 
Talked with at least one of my parents about my activities 
 
Parental Conflict 
 
-How many times in the last week did the following things happen between you and at 
least one of your parents: 
 
We got angry at each other 
We argued during dinner 
We had a big argument about a little thing 
One of us got so mad we hit the other person 
I got my way by getting angry 
 
Prosocial Peer Affiliation 
 
-In the last three months, did your friends: 
 
Complete assignments or homework on time 
Participate in sports or organized activities (such as drama, music, religious group) 
Help with chores around their house  
Cooperate with their parent(s) 
Cooperate with their teachers 
Attend church or religious activities 
Act friendly and helpful to family members 
Follow through with a plan or goal 
Feel confident and proud of an accomplishment 
Feel positive about going to school 
 
193 
 
Antisocial Peer Affiliation 
 
-Over the last three months . . . 
 
How many of our friends have cheated on school tests? 
How many of your friends have ruined or damaged something on purpose that did not 
belong to them? 
 
194 
 
BIBLIOGRAPHY 
Achenbach, T. M., McConaughy, S. H., & Howell, C. T. (1987).  Child/adolescent 
behavioral and emotional problems:  Implications of cross-informant correlations 
for situational specificity.  Psychological Bulletin, 101(2), 213-232. 
 
Ahadi, S. A., Rothbart, M. K., & Ye, R. (1993).  Children's temperament in the US and 
China: Similarities and differences.  European Journal of Personality, 7, 359-377. 
 
Allen, J. P., Moore, C., Kuperminc, G., & Bell, K. (1998).  Attachment and adolescent 
psychosocial functioning.  Child Development, 69(5), 1406-1419. 
 
Angold, A., Costello, E. J., Erkanli, A., & Worthman, C. M. (1999).  Pubertal changes in 
hormone levels and depression in girls.  Psychological Medicine, 29, 1043-1053. 
 
Angold, A., Costello, E. J., & Worthman, C. M. (1998).  Puberty and depression: The 
roles of age, pubertal status, and pubertal timing.  Psychological Medicine, 28, 
51-61. 
 
Arnett, J. J. (1996).  Sensation seeking, aggressiveness, and adolescent reckless behavior.  
Personality & Individual Differences, 20(6). 
 
Arnett, J. J. (1999).  Adolescent storm and stress, reconsidered.  American Psychologist, 
54(5), 317-326. 
 
Ary, D. V., Duncan, T. E., Biglan, A., Metzler, C. W., Noell, J. W., & Smolkowski, K. 
(1999).  Development of adolescent problem behavior.  Journal of Abnormal 
Child Psychology, 27(2), 141-150. 
 
Bates, J. E., Pettit, G. S., & Dodge, K. A. (1995).  Family and child factors in stability 
and change in children's aggressiveness in elementary school.  In J. McCord, 
(Ed), Coercion and punishment in long-term perspectives (pp. 124-138).  New 
York: Cambridge University Press. 
 
Bates, J. E., Pettit, G. S., Dodge, K. A., & Ridge, B. (1998).  Interaction of 
temperamental resistance to control and restrictive parenting in the development 
of externalizing behavior.  Developmental Psychology, 34, 982-995. 
 
Bebbington, P. E., Dunn, G., Jenkins, R., Lewis, G., Brugha, T., Farrell, M., & Meltzer, 
H. (1998).  The influence of age and sex on the prevalence of depressive 
conditions: Report from the National Survey of Psychiatric Morbidity.  
Psychological Medicine, 28, 9-19. 
 
195 
 
Botvinick, M., Nystrom, L. E., Fissell, K., Carter, C. S., & Cohen, J. D. (1999).  Conflict 
monitoring versus selection-for-action in anterior cingulate cortex.  Nature, 
402(6758), 179-181. 
 
Brooks-Gunn, J., Graber, J. A., & Paikoff, R. L. (1994).  Studying links between 
hormones and negative affect: Models and measures.  Journal of Research on 
Adolescence, 4(4), 469-486. 
 
Brooks-Gunn, J., & Warren, M. P. (1989).  Biological and social contributions to 
negative affect in young adolescent girls.  Child Development, 60(1), 40-55. 
 
Brooks-Gunn, J., Warren, M. P., Rosso, J., & Gargiulo, J. (1987).  Validity of self-report 
measures of girls' pubertal status.  Child Development, 58, 829-841. 
 
Buchanan, C. M., Eccles, J. S., & Becker, J. B. (1992).  Are adolescents the victims of 
raging hormones?  Evidence for activational effects of hormones on moods and 
behavior at adolescence.  Psychological Bulletin, 111(1), 62-107. 
 
Buchanan, C. M., & Holmbeck, G. N. (1998).  Measuring beliefs about adolescent 
personality and behavior.  Journal of Youth & Adolescence, 27(5) 607-627. 
 
Bush, G., Frazier, J. A., Rauch, S. L., Seidman, L. J., Whalen, P. J., Jenike, M. A., Rosen, 
B. R., & Biederman, J. (1999).  Anterior cingulate cortex dysfunction in attention-
deficit/hyperactivity disorder revealed by fMRI and the Counting Stroop.  
Biological Psychiatry, 45(12), 1542-1552. 
 
Bush, G., Luu, P., & Posner, M.I. (2000).  Cognitive and emotional influences in the 
anterior cingulate cortex.  Trends in Cognitive Science, 4(6), 215-222. 
 
Bush, G., Whalen, P. J., Rosen, B. R., Jenike, M. A., McInerney, S. C., & Rauch, S. L. 
(1998).  The counting Stroop: An interference task specialized for functional 
neuroimaging - validation study with functional MRI.  Human Brain Mapping, 6, 
270-282. 
 
Capaldi, D. M., & Rothbart, M. K. (1992).  Development and validation of an early 
adolescent temperament measure.  Journal of Early Adolescence, 12, 153-173. 
 
Carskadon, M. A., & Acebo, C. (1993).  A self-administered rating scale for pubertal 
development.  Journal of Adolescent Health, 24, 190-195. 
 
Caspi, A., Henry, B., McGee, R. O., Moffitt, T. E., & et al. (1995).  Temperamental 
origins of child and adolescent behavior problems: From age three to fifteen.  
Child Development, 66(1), 55-68. 
 
196 
 
Cloninger, C. R. (1987).  A systematic method for clinical description and classification 
of personality variants: A proposal.  Archives of General Psychiatry, 44(6), 573-
588. 
 
Collins, W. A., Maccoby, E. E., Steinberg, L., Hetherington, E. M., & Bornstein, M. 
(2000).  Contemporary research on parenting: The case for nature and nurture.  
American Psychologist, 218-231. 
 
Collins, W. A., Maccoby, E. E., Steinberg, L., Hetherington, E. M., & Bornstein, M. H. 
(2001).  Toward nature WITH nurture.  American Psychologist, 56(2), 171-173. 
 
Cyranowski, J. M., Frank, E., Young, E., & Shear, K. (2000).  Adolescent onset of the 
gender difference in lifetime rates of major depression.  Archives of General 
Psychiatry, 57(1), 21-27. 
 
Dahl, R. E. (2001).  Affect regulation, brain development, and behavioral/emotional 
health in adolescence.  Paper presented at the Society for Research in Child 
Development, Minneapolis, MN. 
 
Darling, N., & Cohen, C. L. (2002).  Romantic anxiety and avoidance and middle 
adolescents’ depressive symptoms.  Paper presented at the Society for Research 
on Adolescence, New Orleans, LA.  
 
Davies, P. L., & Rose, J. D. (1999).  Assessment of cognitive development in adolescents 
by means of neuropsychological tasks.  Developmental Neuropsychology, 15(2), 
227-248. 
 
Davila, J., Steinberg, S.J., Kachadourian, L., Cobb, R., Fincham, F. (2002).  Early 
romantic experiences and depressive symptoms:  Emerging depressogenic 
patterns.  Paper presented at the Society for Research on Adolescence, New 
Orleans, LA. 
 
Deater-Deckard, K. (2000).  Parenting and child behavioral adjustment in early 
childhood: A quantitative genetic approach to studying family processes and child 
development. Child Development, 71, 468-484.  
Derryberry, D., & Rothbart, M. K. (1997).  Reactive and effortful processes in the 
organization of temperament.  Development & Psychopathology, 9(4), 633-652. 
 
Dishion, T. J., & Bullock, B. M. (2002).  Parenting and adolescent problem behavior: An 
ecological analysis of the nurturance hypothesis.  In J. G. Borkowski, S. L. 
Ramey, et al. (Eds.); Parenting and the child's world: Influences on academic, 
intellectual, and social-emotional development. Monographs in parenting, (pp. 
231-249).  Mahwah, NJ: Erlbaum. 
 
197 
 
Dishion, T. J., Capaldi, D., Spracklen, K. M., & Li, F. (1995).  Peer ecology of male 
adolescent drug use.  Development & Psychopathology, 7(4), 803-824. 
 
Dishion, T. J., Spracklen, K. M., Andrews, D. W., & Patterson, G. R. (1996).  Deviancy 
training in male adolescents friendships.  Behavior Therapy, 27(3), 373-390. 
 
Doost, H. T. N., Moradi, A. R., Taghavi, M. R., Yule, W., & Dalgleish, T. (1999).  The 
development of a corpus of emotional words produced by children and 
adolescents.  Personality & Individual Differences, 27, 433-451. 
 
Doost, H. T. N., Taghavi, M. R., Moradi, A. R., Yule, W., & Dalgleish, T. (1997).  The 
performance of clinically depressed children and adolescents on the modified 
Stroop paradigm.  Personality & Individual Differences, 23(5), 753-759. 
 
Downey, G., & Aduk O. (2002).  Rejection sensitivity and depressive symptoms in young 
women following a romantic rejection.  Paper presented at the Society for 
Research on Adolescence, New Orleans, LA. 
 
Drigotas, S. M., & Udry, J. R. (1993).  Biosocial models of adolescent problem behavior: 
Extension to panel design.  Social Biology, 40(1-2), 1-7. 
 
Eisenberg, N. (2000).  Emotion, regulation, and moral development.  Annual Review of 
Psychology, 51, 665-697. 
 
Eisenberg, N., Cumberland, A., Spinrad, T. L., Fabes, R. A., Shepard, S. A., Reiser, M., 
Murphy, B. C., Losoya, S. H., & Guthrie, I. K. (2001).  The relations of regulation 
and emotionality to children's externalizing and internalizing problem behavior.  
Child Development, 72(4), 1112-1134. 
 
Ellis, L. K., & Rothbart, M. K. (2002).  Revision of the Early Adolescent Temperament 
Questionnaire.  Manuscript in preparation. 
 
Fabes, R. A., Carlo, G., Kupanoff, K., & Laible, D. (1999).  Early adolescence and 
prosocial/moral behavior I: The role of individual processes.  Journal of Early 
Adolescence, 19(1), 5-16. 
 
Fan, J., McCandliss, B. D., Sommer, T., Raz, M., & Posner, M. I. (2002).  Testing the 
efficiency and independence of attentional networks.  Journal of Cognitive 
Neuroscience, 14(3), 340-347. 
 
Feingold, A. (1994).  Gender differences in personality: A meta-analysis.  Psychological 
Bulletin, 116(3), 429-456. 
 
198 
 
Finkelstein, J. W., von Eye, A., & Preece, M. A. (1994).  The relationship between 
aggressive behavior and puberty in normal adolescents: A longitudinal study.  
Journal of Adolescent Health, 15, 319-326. 
 
Frank, E., & Young, E. (2000).  Pubertal changes and adolescent challenges: Why do 
rates of depression rise precipitously for girls between ages 10 and 15 years?  In 
E. Frank (Ed.), Gender and Its Effects on Psychopathology (pp. 85-102).  
Washington, DC: American Psychiatric Press, Inc. 
 
Ge, X., Conger, R. D., & Elder, G. H. (2001).  Pubertal transition, stressful life events, 
and the emergence of gender differences in adolescent depressive symptoms.  
Developmental Psychology, 37(3), 404-417. 
 
Ge, X., Conger, R. D., & Elder, G. H. J. R. (2001).  The relation between puberty and 
psychological distress in adolescent boys.  Journal of Research on Adolescence, 
11, 49-70. 
 
Gerardi-Caulton, G. (2000).  Sensitivity to spatial conflict and the development of self-
regulation in children 24-36 months of age.  Developmental Science, 3(4), UK: 
Blackwell Publishers. 
 
Giancola, P. R. (2000).  Temperament and antisocial behavior in preadolescent boys with 
or without a family history of a substance use disorder.  Psychology of Addictive 
Behaviors, 14(1). 
 
Giancola, P. R., Mezzich, A. C., & Tarter, R. E. (1998).  Executive cognitive functioning, 
temperament, and antisocial behavior in conduct-disordered adolescent females.  
Journal of Abnormal Psychology, 107(4), 629-641. 
 
Giancola, P. R., Moss, H. B., Martin, C. S., Kirisci, L., & Tarter, R. E. (1996).  Executive 
cognitive functioning predicts reactive aggression in boys at high risk for 
substance abuse: A prospective study.  Alcoholism: Clinical & Experimental 
Research, 20(4), 740-744. 
 
Giancola, P. R., & Tarter, R. E. (1998).  Executive cognitive functioning and risk for 
substance abuse.  Psychological Science, 10(3), 203-205. 
 
Giancola, P. R., & Zeichner, A. (1994).  Neuropsychological performance on tests of 
frontal-lobe functioning and aggressive behavior in men.  Journal of Abnormal 
Psychology, 103(4), 832-835. 
 
Giedd, J. N., Vaituzis, A. C., Hamburger, S. D., Lange, N., Rajapakse, J. C., Kaysen, D., 
Vauss, Y. C., & Rapoport, J. L. (1996).  Quantitative MRI of the temporal lobe, 
amygdala, and hippocampus in normal human development: Ages 4-18 years.  
The Journal of Comparative Neurology, 366(22), 223-230. 
199 
 
 
Gottfredson, M. R., & Hirschi, T. (1990).  A general theory of crime.  Stanford 
University Press. 
 
Graber, J. A., Lewinsohn, P. M., Seeley, J. R., & Brooks-Gunn, J. (1997).  Is 
psychopathology associated with the timing of pubertal development?  Journal of 
the American Academy of Child & Adolescent Psychiatry, 36(12), 1768-1776. 
 
Graber, J. A., Petersen, A. C., & Brooks-Gunn, J. (1996).  Pubertal Processes: Methods, 
measures, and models.  In J. A. Graber & J. Brooks-Gunn & A. C. Petersen 
(Eds.), Transitions through adolescence:  Interpersonal domains and contexts.  
(pp. 23-54). Mahwah: Lawrence Erlbaum Associates, Inc. 
 
Gray, J. A. (1987).  The neuropsychology of emotion and personality.  In S. D. Stahl, 
(Ed); Cognitive Neurochemistry, (pp. 171-190).  New York: Oxford University 
Press. 
 
Hall, G. S. (1904).  Adolescence.  New York: Appleton. 
 
Hankin, B. L., Abramson, L. Y., Moffitt, T. E., Silva, P. A., McGee, R., & Angell, K. E. 
(1998).  Development of depression from preadolescence to young adulthood: 
Emerging gender differences in a 10-year longitudinal study.  Journal of 
Abnormal Psychology, 107(1), 128-140. 
 
Harper, M.S., Welsh, D., & Wooddy T. (2002).  Silencing the self:  Depressive symptoms 
and the loss of self in adolescent romantic relationships.  Paper presented at the 
Society for Research on Adolescence, New Orleans, LA. 
 
Insel, T. R., Winslow, J. T., & Wang, Z. C. (1995).  Oxytocin and the molecular basis of 
monogamy.  Advances in Experimental Medical Biology, 395, 227-234. 
 
Irwin, Jr. C. E. (1993).  Adolescence and risk taking:  How are they related?  In N. J. 
Bell, & R. W. Bell, (Eds.), Adolescent Risk Taking, (pp. 7-28).  Newbury Park, 
CA: Sage Publications. 
 
Jensen, P. S., Rubio-Stipec, M., Canino, G., Bird, H., Dulcan, M., Schwab-Stone, M. E., 
& Lahey, B. B. (1999).  Parent and child contributions to diagnosis of mental 
disorder: Are both informants always necessary?  Journal of the American 
Academy of Child & Adolescent Psychiatry, 38(12), 1569-1579. 
 
Jernigan, T. L., Trauner, D. A., Hesselink, J. R., & Tallal, P. A. (1991).  Maturation of 
human cerebrum observed in vivo during adolescence.  Brain, (114), 2037-2049. 
 
Kandel, D. B., & Raveis, V. H. (1989).  Cessation of illicit drug use in young adulthood.  
Archives of General Psychiatry, 46(2) 109-116. 
200 
 
 
Klingberg, T., Vaidya, C. J., Gabrieli, J. D. E., Moseley, M. E., & Hedehus, M. (1999).  
Myelination and organization of the frontal white matter in children: A diffusion 
tensor MRI study.  Neuroreport: an International Journal for the Rapid 
Communication of Research in Neuroscience, 10(13), 2817-2821. 
 
Kochanska, G. (1997).  Multiple pathways to conscience for children with different 
temperaments: From toddlerhood to age 5.  Developmental Psychology, 33(2), 
228-240. 
 
Kochanska, G., Murray, K., & Harlan, E. (2000).  Effortful control in early childhood:  
Continuity and change, antecedents, and implications for social development.  
Developmental Psychology, 36(1), 220-232. 
 
Kopstein, A., Crum, R. M., Celentano, D. D., & Martin, S. S. (2001).  Sensation seeking 
needs among 8th and 11th graders: Characteristics associated with cigarette and 
marijuana use.  Drug and Alcohol Dependence, 62, 195-203. 
 
Krueger, R. F., Caspi, A., Moffitt, T. E., White, J., Stouthamer-Loeber, M. (1996).  Delay 
of gratification, psychopathology, and personality: Is low self-control specific to 
externalizing problems?  Journal of Personality, 64(1) 107-129. 
 
Laitinen-Krispijn, S., Van der Ende, J., Hazebroek-Kampschreur, A. A. J. M., & 
Verhulst, F. C. (1999).  Pubertal maturation and the development of behavioural 
and emotional problems in early adolescence.  Acta Psychiatrica Scandinavica, 
99(1), 16-25. 
 
Larson, R., & Richards, M. H. (1989).  Introduction: The changing life space of early 
adolescence.  Journal of Youth & Adolescence, 18(6), 501-509. 
 
Larson, R., & Richards, M. H. (1994).  Divergent realities: The emotional lives of 
mothers, fathers, and adolescents.  New York: Basic Books, Inc. 
 
Larson, R. & Richards, M. H. (1996).  Divergent Realities:  The emotional lives of 
mothers, fathers, and adolescents.  New York: Basic Books, Inc. 
 
Lee, C. L., & Bates, J. E. (1985).  Mother-child interaction at age two years and 
perceived difficult temperament.  Child Development, 56, 1314-1325. 
 
Levin, H. S., Culhane, K. A., Hartmann, J., Evankovich, K., & et al. (1991).  
Developmental changes in performance on tests of purported frontal lobe 
functioning.  Developmental Neuropsychology, 7(3), 377-395. 
 
201 
 
Lewis, D. A. (1997).  Development of the prefrontal cortex during adolescence: Insights 
into vulnerable neural circuits in schizophrenia.  Neuropsychopharmacology, 
16(6), 385-398. 
 
Mattsson, A., Schalling, D., Olweus, D., Low, H., & Svensson, J. (1980).  Plasma 
testosterone, aggressive behavior, and personality dimensions in young male 
delinquents.  Journal of the American Academy of Child Psychiatry, 19, 476-490. 
 
McCrae, R. R., Costa, P. T., Ostendorf, F., Angleitner, A., Hrebickova, M., & Avia, M. 
(2000).  Nature over nurture: Temperament, personality, and life span 
development.  Journal of Personality and Social Psychology, 78(1), 173-186. 
 
McNeil, D.W., Tucker, P., Miranda Jr, R., Lewin, M.R., & Nordgran, J.C. (1999).  
Response to depression and anxiety Stroop stimuli in posttraumatic stress 
disorder, obsessive-compulsive disorder, and major depressive disorder.  Journal 
of Nervous & Mental Disease, 187(8), 512-516. 
 
Mead, M. (1928).  Coming of age in Samoa.  New York: Morrow. 
 
Metzler, C. W., Biglan, A., Ary, D. V., & Fuzhong, L. (1998).  The stability and validity 
of early adolescents' reports of parenting constructs.  Journal of Family 
Psychology, 12(4), 600-619. 
 
Mezzacappa, E., Kindlon, D., & Earls, F. (1999).  Relations of age to cognitive and 
motivational elements of impulse control in boys with and without externalizing 
behavior problems.  Journal of Abnormal Child Psychology, 27(6), 473-483. 
 
Moffitt, T. E. (1993).  The neuropsychology of conduct disorder.  Development & 
Psychopathology, 5(1-2), 135-151. 
 
Moffit, T. E., Caspi, A., Dickson, N., Silva, P., & Stanton, W. (1996).  Childhood-onset 
versus adolescent-onset antisocial conduct problems in males: Natural history 
from ages 3 to 18 years.  Development & Psychopathology, 8(2), 399-424. 
 
Moffitt, T. E., Henry, B. (1989).  Neuropsychological assessment of executive functions 
in self-reported delinquents.  Development & Psychopathology. 1(2) 105-118. 
 
Moradi, A. R., Taghavi, M. R., Doost, H. T. N., Yule, W., & Dalgleish, T. (1999).  
Performance of children and adolescents with PTSD on the Stroop colour-naming 
task.  Psychological Medicine, 29, 415-419. 
 
Nolen-Hoeksema, S., & Girgus, J. (1994).  The emergence of gender differences in 
depression during adolescence.  Psychological Bulletin, 115(3), 424-443. 
 
202 
 
Norman, D. A., & Shallice, T. (1986).  Attention to action:  Willed and automatic control 
of behavior.  In R.J. Davidson, G.E. Schwartz, and D. Shapiro (Eds.), 
Consciousness and Self-Regulation.  (Vol. 4, pp.1-18).  New York: Plenum Press. 
 
Nottelmann, E. D., Susman, E. J., Blue, J. H., Inoff-Germain, G., Dorn, L. D., Loriaux, 
D. L., & Cutler, G. P. (1987).  Gonadal and adrenal hormone correlates of 
adjustment in early adolescence.  In R. M. Lerner & T. T. Foch, (Eds.), 
Biological-psychosocial interactions in early adolescence, (pp.303-323).  
Hillsdale, NJ: Erlbaum. 
 
Paikoff, R. L., Brooks-Gunn, J., & Warren, M. P. (1991).  Predictive effects of hormonal 
change on affective expression in adolescent females over the course of one year.  
Journal of Youth and Adolescence, 20, 191-214. 
 
Patterson, G. R. (1982).  A social learning approach:  III Coercive family process.  
Eugene, OR: Castalia. 
 
Patterson, G. R., DeBaryshe, B. D., & Ramsey, E. (1989).  A developmental perspective 
on antisocial behavior.  American Psychologist, 44(329-335). 
 
Patterson, G. R., Reid, J., & Dishion, T. J. (1992).  A social interactional approach IV:  
Antisocial boys.  Eugene, OR: Castalia. 
 
Paus, T., Zijdenbos, A., Worsley, K., Collins, D. L., Blumenthal, J., Giedd, J. N., 
Rapoport, J. L., & Evans, A. C. (1999).  Structural maturation of neural pathways 
in children and adolescents: In vivo study.  Science, 283(5409), 1908-1911. 
 
Petersen, A. C., Compas, B. E., Brooks-Gunn, J., Stemmler, M., & et al. (1993).  
Depression in adolescence.  American Psychologist, 48(2). 
 
Petersen, A. C., Crockett, L., & Richards, M. (1988).  A self-report measure of pubertal 
status:  Reliability, validity, and initial norms.  Journal of Youth & Adolescence, 
17, 117-133. 
 
Petersen, A. C., Crockett, L., Richards, M., & Boxer, A. (1988).  A self-report measure of 
pubertal status: Reliability, validity, and initial norms.  Journal of Youth & 
Adolescence, 17(2), 117-133. 
 
Plant, T. M. (2001).  Neurobiological bases underlying the control of the onset of puberty 
in the rhesus monkey: A representative higher primate.  Frontiers in 
neuroendocrinology, 22(2), 107-140. 
 
Posner, M. I., & DiGirolamo, G. J. (1998).  Executive attention: Conflict, target 
detection, and cognitive control.  In R. Parasuraman (Ed.), The attentive brain, 
(pp. 401-423).  Cambridge, MA: The MIT Press. 
203 
 
 
Posner, M. I., & Petersen, S. E. (1990).  The attention system of the human brain.  Annual 
Review of Neuroscience, 13, 25-42. 
 
Posner, M. I., & Rothbart, M. K. (1998).  Attention, self-regulation, and consciousness.  
Phil. Trans. of the Royal Society of London, 353, 1915-1927. 
 
Posner, M. I., & Rothbart. M. K. (2000).  Developing mechanisms of self-regulation.  
Development and Psychopathology, 12(3), 427-441.   
 
Putnam, S. P., Ellis, L. K., & Rothbart, M. K. (2001).  The structure of temperament.  In 
A. Eliasz & A. Angleitner (Eds.), Advances in Research on Temperament (pp. 
165-182).  Lengerich: Pabst Scientific Publishers. 
 
Putnam, S. P., Sanson, A. V., & Rothbart, M. K. (in press).  Child Temperament and 
parenting.  In M. Bornstein (Ed.), Handbook of Parenting (Vol. 2nd).  NJ: 
Erlbaum. 
 
Richards, A., Richards, L., & McGeeney, A. (2000).  Anxiety-related Stroop interference 
in adolescents.  Journal of General Psychology, 127(3), 327-333. 
 
Richards, M. H., Crowe, P. A., Larson, R., & Swarr, A. (1998).  Developmental patterns 
and gender differences in the experience of peer companionship during 
adolescence.  Child Development, 69(1), 154-163. 
 
Romero, E., Luengo, M. A., & Sobrol, J. (2001).  Personality and antisocial behaviour: 
Study of temperamental dimensions.  Personality & Individual Differences, 31, 
329-348. 
 
Rothbart, M. K., Ahadi, S. A., & Evans, D. E. (2000).  Temperament and personality: 
Origins and outcomes.  Journal of Personality and Social Psychology, 78(1), 122-
135. 
 
Rothbart, M. K, Ahadi, S. A., & Hershey, K. S. (1994).  Temperament and social 
behavior in childhood.  Merrill-Palmer Quarterly, 40, 21-39.   
 
Rothbart, M. K., & Bates, J. E. (1998).  Temperament.  In N. Eisenberg (Ed.), Vol. 3:  
Social, Emotional and Personality Development (5th ed., Vol. 3, pp. 105-176).  
New York: J. Wiley. 
 
Rothbart, M. K., Derryberry, D., & Posner, M. I. (1994).  A psychobiological approach to 
the development of temperament.  In J. E. Bates & T. D. Wachs (Eds.), 
Temperament: Individual differences at the interface of biology and behavior.  
APA science volumes. (pp. 83-116).  Washington, DC: American Psychological 
Association. 
204 
 
 
Rothbart, M. K., Ellis, L. K., Rueda, M. R., & Posner, M. I. (2002).  Developing 
mechanisms of temperamental effortful control.  Manuscript in submission. 
 
Rothbart, M. K., Posner, M. I., & Hershey, K. (1995).  Temperament, attention, and 
developmental psychopathology.  In D. Cicchetti & D. J. Cohen (Eds.), 
Developmental Psychopathology, Vol. 1:  Theory and Methods (pp. 315-340).  
New York: Wiley. 
 
Rueda, M. R. C., & Posner M. I. (2002).  Measuring attentional networks in children.  
Manuscription in submission. 
 
Rutter, M., Graham, P., Chadwick, F., & Yule, W. (1976).  Adolescent turmoil: Fact or 
fiction?  Journal of Adolescent Psychology and Psychiatry, 17, 35-56. 
 
Scerbo, A. S., & Kolko, D. J. (1994).  Salivary testosterone and cortisol in disruptive 
children: Relationship to aggressive, hyperactive, and internalizing behaviors.  
Journal of the American Academy of Child & Adolescent Psychiatry, 33(8), 1174-
1184. 
 
Sowell, E. R., & Jernigan, T. L. (1998).  Further MRI evidence of late brain maturation: 
Limbic volume increases and changing asymmetries during childhood and 
adolescence.  Developmental Neuropsychology, 14(4), 599-617. 
 
Sowell, E. R., Thompson, P. M., Holmes, C. J., Jernigan, T. L., & Toga, A. W. (1999).  
In vivo evidence for post-adolescent brain maturation in frontal and striatal 
regions.  Nature Neuroscience, 2(10), 859-861. 
 
Spear, L. P. (2000).  The adolescent brain and age-related behavioral manifestations.  
Neuroscience & Biobehavioral Reviews, 24, 417-463. 
 
Stattin, H., & Magnusson, D. (1990).  Pubertal maturation in female development.  
Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. 
 
Steinberg, L. (2001).  We know some things: Parent-adolescent relationships in retrospect 
and prospect.  Journal of Research on Adolescence, 11(1), 1-19. 
 
Susman, E. J. (1997).  Modeling developmental complexity in adolescence: Hormones 
and behavior in context.  Journal of Research on Adolescence, 7(3), 283-306. 
 
Susman, E. J., Inoff-Germain, G., Nottelmann, E. D., Loriaux, D. L, Cutler, G. B., & 
Chrousos, G. P. (1987).  Hormones, emotional dispositions, and aggressive 
attributes in young adolescents.  Child Development, 58, 1114-1134. 
 
205 
 
Taylor, S. E., Klein, L. C., Lewis, B. P., Gruenewald, T. L., Gurung, R. A. R., & 
Updegraff, J. A. (2000).  Biobehavioral responses to stress in females: Tend-and-
befriend, not fight-or-flight.  Psychological Review, 107(3), 411-429. 
 
Terasawa, E., & Fernandez, D. L. (2001).  Neurobiological mechanisms of the onset of 
puberty in primates.  Endocrine Reviews, 22(1), 111. 
 
Thomas, A. & Chess, S. (1977).  Temperament and Development.  New York: 
Brunner/Mazel. 
 
Thompson, P. M., Giedd, J. N., Woods, R. P., MacDonald, D., Evans, A. C., & Toga, A. 
W. (2000).  Growth patterns in the developing brain detected by using continuum 
mechanical tensor maps.  Nature, 404(3), 189-192. 
 
Tremblay, R. E., Schaal, B., Boulerice, B., Arseneault, L., & Soussignan, R. (1998).  
Testosterone, physical aggression, dominance, and physical development in early 
adolescence.  International Journal of Behavioural Development, 22(4), 753. 
 
Tubman, J. G. & Windle, M. (1995).  Continuity of difficult temperament in adolescence:  
Relations with depression, life events, family support, and substance use across a 
one-year period.  Journal of Youth and Adolescence, 24, 133-153. 
 
Turner, R. A., Altemus, M., Enos, T., Cooper, B., & McGuinness, T. (1999).  Preliminary 
research on plasma oxytocin in normal cycling women: Investigating emotion and 
interpersonal distress.  Psychiatry: Interpersonal & Biological Processes, 62(2), 
97-113. 
 
Udry, J. R., & Talbert, L. M. (1988).  Sex hormone effects on personality at puberty.  
Journal of Personality and Social Psychology, 54, 291-295. 
 
Uvnas-Moberg, K. (1999).  Oxytocin may mediate the benefits of positive social 
interaction and emotions.  Psychoneuroendocrinology, 23(8), 819-835. 
 
van den Boom, D. C. (1994).  The influence of temperament and mothering on 
attachment and exploration: An experimental manipulation of sensitive 
responsiveness among lower-class mothers with irritable infants.  Child 
Development, 65(5), 1457-1477. 
 
Walker, E. F. (2002).  Adolescent neurodevelopment and psychopathology.  Current 
Directions in Psychological Science, 11(1), 24-28. 
 
Whalen, P. J., Bush, G., McNally, R. J., Wilhelm, S., McInerney, S. C., Jenike, M. A., & 
Rauch, S. L. (1998).  The emotional counting Stroop paradigm: A functional 
magnetic resonance imaging probe of the anterior cingulate affective division.  
Biological Psychiatry, 44, 1219-1228. 
206 
 
 
White, J. L., Moffitt, T. E., Caspi, A., Bartusch, D. J., & et al. (1994).  Measuring 
impulsivity and examining its relationship to delinquency.  Journal of Abnormal 
Psychology, 103(2), 192-205. 
 
Williams, B. R., Ponesse, J. S., Schachar, R. J., Logan, G. D., & Tannock, R. (1999).  
Development of inhibitory control across the life span.  Developmental 
Psychology, 35(1). 
 
Williams, J. M. G., Matthews, A., & MacLeod, C. (1996).  The emotional Stroop task 
and psychopathology.  Psychological Bulletin, 120(1), 3-24.  
 
Wills, T. A., Cleary, S., Filer, M., Shinar, O., Mariani, J., & Spera, K. (2001).  
Temperament related to early-onset substance use: Test of a developmental 
model.  Prevention Science, 2(3), 145-163. 
 
Wills, T. A., Sandy, J. M., & Yaeger, A. (2001).  Time perspective and early-onset 
substance use: A model based on stress-coping theory.  Psychology of Addictive 
Behaviors, 15(2), 118-125. 
 
Wills, T. A., Sandy, J. M., & Yaeger, A. (in press).  Moderators of the relationship 
between substance use level and problems: Test of a self-regulation model in 
middle adolescence.  Journal of Abnormal Psychology. 
 
Wills, T. A., Sandy, J. M., Yaeger, A., & Shinar, O. (2001).  Family risk factors and 
adolescent substance use: Moderation effects for Temperament Dimensions.  
Developmental Psychology, 37(3), 283-297. 
 
Wills, T. A., Windle, M. & Cleary, S. D.  (1998).  Temperament and novelty seeking in 
adolescent substance use:  Convergence of dimensions of temperament with 
constructs from Cloninger’s theory.  Journal of Personality and Social 
Psychology, 74, 387-406. 
 
Windle, M., Hooker, K., Lenerz, K. East, P. L., Lerner, J. V., & Lerner, R. M. (1986).  
Temperament, perceived competence, and depression in early and late 
adolescents.  Developmental Psychology, 22, 384-392. 
 
Zuckerman, M. (1984).  Sensation seeking: A comparative approach to a human trait.  
Behavioral and Brain Sciences, 7, 413-471. 
 
Zuckerman, M. (1988).  Sensation seeking: A comparative approach to a human trait.  
Behavioral and Brain Sciences, 7, 413-471.