MEASURING FACULTY ATTITUDES AND PERCEPTIONS TOWARD DISABLITY
AT A FOUR-YEAR UNIVERSITY: A VALIDITY STUDY
by
ALLISON LOMBARDI
A DISSERTATION
Presented to the Department of Special Education
and Clinical Sciences
and the Graduate School of the University of Oregon
in partial fulfillment of the requirements
for the degree of
Doctor ofPhilosophy
March 2010
11
University of Oregon Graduate School
Confirmation of Approval and Acceptance of Dissertation prepared by:
Allison Lombardi
Title:
"Measuring Faculty Attitudes and Perceptions toward Disability at a Four-year University: A
Validity Study"
This dissertation has been accepted and approved in partial fulfillment of the requirements for
the Doctor of Philosophy degree in the Department of Special Education and Clinical Sciences
by:
K Brigid Flannery, Chairperson, Special Education and Clinical Sciences
Deborah Olson, Member, Special Education and Clinical Sciences
Joe Stevens, Member, Educational Methodology, Policy, and Leadership
Douglas Blandy, Outside Member, Arts and Administration
and Richard Linton, Vice President for Research and Graduate Studies/Dean of the Graduate
School for the University of Oregon.
March 20, 2010
Original approval signatures are on file with the Graduate School and the University of Oregon
Libraries.
© 2010 Allison Renee Lombardi
111
Allison Renee Lombardi
An Abstract of the Dissertation of
for the degree of
iv
Doctor of Philosophy
in the Department of Special Education and Clinical Sciences
to be taken March 2010
Title: MEASURING FACULTY ATTITUDES AND PERCEPTIONS TOWARD
DISABILITY AT A FOUR-YEAR UNIVERSITY: A VALIDITY STUDY
Approved: _
Dr. K. Brigid Flannery
The primary purpose of this study was to develop and field test the ExCEL survey
instrument, intended to measure faculty attitudes toward and perceptions about students
with disabilities at four-year colleges and universities. A secondary purpose was to
examine how demographic characteristics of the sample and prior disability-focused
training experiences influenced these attitudes and perceptions. A cross validation study
was conducted using exploratory and confirmatory factor analysis. The exploratory factor
analysis (EFA) resulted in an eight factor structure: Knowledge of Disability Law,
Fairness, Invitation to Disclose, Willingness to Invest Time, Accessibility of Course
Materials, Willingness to Adjust Course Assignments, Satisfaction with Campus
Resources, and Willingness to Make Testing Accommodations. The confirmatory factor
analysis (CFA) was intended to confirm the factor structure but resulted in an inadmissible
vsolution, potentially due to model misspecification. These findings show initial reliability
and validity of the ExCEL survey and suggest item development as a crucial next step in
the development process. Follow-up analyses showed minimal differences between faculty
according to gender but significant differences according to faculty rank (e.g., tenured,
tenure-track, and non-tenure) and department. In addition, group differences emerged
according to prior disability-focused training, where faculty who had previous disability-
focused training had greater positive attitudes and perceptions toward students with
disabilities than did faculty without such training. These results will inform the field about
important constructs to consider in instrument development for measuring faculty attitudes
and perceptions toward disability, including knowledge of legal mandates and
accommodations and lmowledge of and willingness to adjust instruction according to the
tenets of Universal Design. As well, the findings will inform four-year college and
university administrations when planning targeted interventions for faculty members in
disability awareness training.
VI
CURRICULUM VITAE
NAME OF AUTHOR: Allison R. Lombardi
PLACE OF BIRTH: Pittsburgh, Pennsylvania
DATE OF BIRTH: November 1, 1977
GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED:
University of Oregon, Eugene
University of California, Berkeley
DEGREES AWARDED:
Doctor ofPhilosophy, Special Education and Clinical Sciences, 2010, University
of Oregon
Master of Arts, Education, 2003, University of California, Berkeley
Bachelor of Arts, 2000, University of California, Berkeley
AREAS OF SPECIAL INTEREST:
Secondary to Postsecondary Transition
Quantitative Measurement
PROFESSIONAL EXPERIENCE:
Research assistant, Secondary Special Education and Transition, University of
Oregon, Eugene, 2008-present
Research specialist, Educational Policy Improvement Center, Eugene, Oregon,
2007-present
Learning specialist, Services for Student Athletes, University of Oregon, Eugene,
2004-2007
VB
GRANTS, AWARDS AND HONORS:
Demonstration Projects to Ensure Students with Disabilities Receive a Quality
Higher Education, Office ofPostsecondary Education, 2008
PUBLICATIONS:
Murray, C., Lombardi, A., & Wren. C. (in press). The effects of disability-focused
training on the attitudes and perceptions of university staff. Remedial and
Special Education.
Murray, C., Lombardi, A., Wren, C. , & Keys, C. (2009). Associations between
prior disability-focused training and disability-related attitudes and
perceptions among university faculty. Learning Disability Quarterly, 32(2),
89-100.
Lombardi, A. (2008). "Student Athlete Learning Issues" chapter in NACADA
Monograph Series #18 Advising Student Athletes: A Collaborative
Approach to Success.
Vlll
TABLE OF CONTENTS
Chapter Page
I. LITERATURE REVIEW 1
Statement of the Problem 1
The Prevalence of Students with Disabilities at Four-year Colleges and 2
Universities .
Systemic Differences between High School and College......... 4
Differences in Legal Rights 4
Instructional Climate.................. 5
Student Perceptions of Faculty Attitudes....................................................... 6
Inclusive Instructional Practices 8
Faculty Attitudes and Perceptions 9
Type of Accommodation Request 12
Faculty Training and Professional Development.. 13
Universal Design......................................................................................... 14
Summary of the Problem 16
Measuring Faculty Attitudes and Perceptions toward Disability........... 18
Establishing Validity...................................................................................... 22
Factor Analysis 23
Research Questions........................................................................................ 25
II. NIETHODS........................................................................................................... 26
Instrument......................................................................................................... 26
Participants........................................................................................................ 28
Procedures......................................................................................................... 29
Sample and Response Rate 30
Data Analyses 33
IX
Chapter Page
III. RESlTLTS 37
Analyses............................................................................................................. 37
Exploratory Factor Analysis 39
Descriptive Statistics and Factor Reliability............ 51
Confirmatory Factor Analysis........................................................................ 53
Group Comparisons 56
IV. DISCUSSION..................................................................................................... 68
Factor Structure 68
Willingness to Accommodate 70
Universal Design............................................................................................ 73
Reliability 75
Validity............................................................................................................... 75
Group Differences.............................................................................................. 77
Limitations..... 82
Implications for Practice..... 84
APPENDICES 87
A. EXCEL SURVEY 87
B. FIRST RECRUITMENT EMAIL TO FACULTY PARTICIPANTS 96
C. LETTER OF CONSENT TO PARTICIPATE: FACULTy......................... 98
D. SECOND EMAIL TO FACULTYPARTICIPANTS 100
E. THIRD EMAIL TO FACULTY PARTICIPANTS....................................... 101
F. FOURTH EMAIL TO FACULTY PARTICIPANTS 102
G. SEClTRITY QUESTIONS..... 103
H. ITEM MEANS AND STANDARD DEVIATIONS..................................... 104
1. ITEM CORRELATION MATRIX. 106
REFERENCES 109
LIST OF TABLES
Table Page
1. Seven Principles of Universal Design 16
2. Sample Characteristics and Response Rates ofFaculty by College
Membership................................................................................................ 31
3. Communalities of ExCEL Survey Items 40
4. Total Variance Explained ofUnrotated Eigenvalues and Oblique
Rotation Eigenvalues.................................................................................. 43
5. Pattern Coefficients Resulting from EFA with PAF and Oblique Rotation.. 45
6. Summary of Factor Intercorrelations 52
7. Descriptive Statistics and Reliability Coefficients for Full ExCEL Survey
and Subscales................ 53
8. Standardized Discriminant Function Coefficients (SDFCs) and Structure
Coefficients (SCs) for Each Subscale by Teaching Status 58
9. Standardized Discriminant Function Coefficients (SDFCs) and Structure
Coefficients (SCs) for Each Subscale by the Interaction of Gender by
Teaching Status.................................................. 60
10. Standardized Discriminant Function Coefficients (SDFCs) and Structure
Coefficients (SCs) for Each Subscale by College Membership 64
11. Standardized Discriminant Function Coefficients (SDFCs) and Structure
Coefficients (SCs) for Each Subscale by Prior Training 67
x
Xl
LIST OF FIGURES
m~re p~
1. Representation of all possible faculty participants. 31
2. Representation of faculty sample included in study..................................... 32
3. CFA model of ExCEL survey instrument..................................................... 55
1CHAPTER I
LITERATURE REVIEW
Statement of the Problem
College student diversity has dramatically increased over recent decades. Shifts
in student demographics continue, resulting notably in an increase in enrollment for
students with disabilities (Scott, McGuire & Shaw, 2003). Over time this population has
continued to increase on four-year college and university campuses nationwide (Mull,
Sitlington, & Alper, 2001; U.S. Department of Education, 1999), as students with
disabilities now represent approximately nine percent of the total college population
(Hom, Peter, & Rooney, 2002). As a result, there are new challenges for postsecondary
students and faculty alike, particularly regarding accommodation requests for students
with either physical or learning disabilities. Despite these challenges, it is important to
increase the number of students with disabilities attending and completing four-year
colleges and universities in order to ensure that all students not only have an equal
opportunity to a quality higher education but are equipped to compete in an advanced and
global economy.
A reported barrier to postsecondary success of students with disabilities is
negative faculty attitudes and perceptions about disability (Leyser, Vogel, Wyland,
Brulle, 1998; Reed, Lund-Lucas & O'Rourke, 2003; Vogel, Leyser, Wyland & Brulle,
1999). In ranking students with and without disabilities according to qualification for
2admission into various postsecondary institutions, more students with disabilities were
less qualified for admittance than their non-disabled counterparts (U.S. Department of
Education, 1999). Because of this, faculty members might assume all students with
disabilities do not meet the minimum standard admission guidelines for the institution
and therefore are not as qualified as their non-disabled peers. This assumption may lead
to discriminatory attitudes and behaviors of faculty members when approached with
accommodation requests, attitudes that directly impact the retention rates of students with
disabilities. Four-year colleges and universities should assess their faculty members'
attitudes and perceptions of disability in order to determine targeted interventions for
professional development training in disability awareness. In order to do this, a reliable
and valid measure must be developed that adequately captures the constructs associated
with faculty attitudes and perceptions toward students with both physical and learning
disabilities.
The prevalence of students with disabilities at four-year colleges and
universities. Although enrollment rates of students with disabilities have increased over
time, discrepancies persist between students with and without disabilities. For example,
students with disabilities are less likely than their non-disabled peers to attend a 4-year
college or university after completing high school. Postsecondary enrollment and
completion rate discrepancies exist for students with both learning and physical
disabilities (U.S. Department of Education, 1999).
Murray, Goldstein, Nourse, and Edgar (2000) found that only 9% of young adults
with learning disabilities had attended four-year colleges or universities five years after
3high school graduation; in comparison, 62% of young adults without learning disabilities
continued their education at the same type of postsecondary institutions. Further, only
2% of the adults with learning disabilities had graduated from four-year postsecondary
institutions a full ten years following high school graduation, whereas 46% of the
nondisabled comparison group had completed undergraduate programs (Murray et aI.,
2000).
More recently, Wagner, Newman, Cameto, Levine, and Garza (2006) found that
7% of students with learning disabilities reported future plans to complete a four-year
degree, and approximately 3 in 10 out-of-school youth with either physical or learning
disabilities have enrolled in some type of postsecondary education, as compared to a 41 %
postsecondary enrollment rate of their non-disabled peers. Further, Wagner, Newman,
Cameto, Levine, and Marder (2007) found that although most high school students with
either physical or learning disabilities reported feeling confident they would earn their
diploma, they felt less confident that they would attend some level of postsecondary
schooling. These findings show that both students with physical and/or learning
disabilities are a minority population on four-year college and university campuses.
There is evidence that students with learning disabilities are less likely to attend
four-year colleges and universities than students with Speech Language Impairments,
Hearing Impairments, Visual Impairments, and Orthopedic Impairments (U. S.
Department of Health and Human Services, 2003; U.S. Department of Labor, 2007;
Wagner et aI., 2006). Although these findings suggest students with physical disabilities
are more likely to pursue and complete a postsecondary degree program than students
4with learning disabilities, regardless of disability type, discrepancies persist between the
postsecondary enrollment and completion rates of students with and without disabilities.
Together, these findings support why students with disabilities continue to be an
underrepresented group at four-year college and university settings.
Systemic differences between high school and college. Due to differences
between IDEA and Section 504 of the Rehabilitation Act of 1973 and the Americans with
Disabilities Act, students with disabilities may experience a rough transition particularly
in the areas oflegal rights, curriculum, and instructor training (Scott et aI., 2003). While
many high school students with disabilities report that they intend to pursue some type of
employment, fewer report that they will pursue postsecondary schooling, particularly at a
4-year college or university (Wagner et aI., 2006). Students' lack of confidence to
continue their education may be the result of two major differences between K-12 and
postsecondary settings: legal rights and instructional climate.
Differences in legal rights. Stark differences in support services between K-12
and postsecondary settings are particularly true for students with learning disabilities.
For example, postsecondary students with disabilities are not legally entitled to
specialized instruction, as they may be at the secondary level. Instead, college students
with disabilities are fully integrated in the general student population, and are expected to
self-advocate for their needs. Further, K-12 curricula are more likely to be formed based
on federal and state standards and policy mandates, and therefore more standardized and
consistent within states. Postsecondary curricula encompass a wide range of information
and objectives, and are not subject to state standards (Scott et aI., 2003). This lack of
5standardization allows college faculty members a vast amount of freedom when
curriculum planning. Too often, this freedom leads faculty members to adopt
instructional methods that are more similar to their own learning preferences, and less
responsive to student needs (Ouellett, 2004). Although differences between the two
systems, and their subsequent legal mandates, obviously exist, systemic differences are
rarely considered when rationalizing the discrepancy in postsecondary achievement
between students with and without disabilities. Typically, blame is placed on the
individual student's lack of preparation rather than acknowledgment of systemic
differences. Further study of the discrepancies between students with disabilities and
their non-disabled peers at four-year college settings is warranted, particularly from a
systems-level perspective.
Instructional climate. Federal and state regulations and laws are not the only
differences between the K-12 and postsecondary systems. It is common for most
postsecondary faculty to have expertise in specialized content areas, not in effective
pedagogical methods (Ouellett, 2004; Scott et aI., 2003). In addition, there are typically
little to no required training programs on disability awareness. While some voluntary
training programs exist, published descriptions of these efforts are less common (Murray,
Wren, Stevens, & Keys, 2009). As a result, faculty members are often unaware of
acceptable accommodation requests and effective instructional practices that can enhance
the learning environment for all students, including students with disabilities
(Burgstahler, Duclos, & Turcotte, 1999; Ouellet, 2004).
6At the same time, effective teaching has become an area of professorial merit. It
is becoming more common for faculty members to undergo perfoffi1ance reviews that
include the quality of instruction delivered to students (Scott et aI., 2003). This shift is
perhaps the result of the increasing perception of college students as consumers of
education. Therefore, it is in the best interest of postsecondary faculty members to
participate in professional development, particularly in the area of disability awareness
and the use of inclusive instructional principles that will benefit all students. However,
professional development interventions are difficult to plan and implement without an
initial assessment of the faculty attitudes and perceptions toward students with disabilities
at a given university.
Student perceptions of faculty attitudes. Students with disabilities enrolled in
colleges and universities have expressed that faculty attitudes toward individuals with
disabilities and their understanding of disabilities has a direct impact on their motivation
and determination to succeed at the postsecondary level (Dowrick, Anderson, Heyer, &
Acosta, 2005; Farone, Hall, & Costello, 1998; Hartman-Hall & Haaga, 2002; Parker,
Embry, Scott, & McGuire, 2003). Research findings from the students' perspective
indicate that some faculty, staff, and administrators lacked information on disability
rights, generally displayed poor attitudes toward students with disabilities, and did not
always show willingness to accommodate (Dowrick et aI., 2005; Farone et aI., 1998).
In a study by Hartman-Hall and Haaga (2002), students were asked to react to
hypothetical scenarios that involved faculty expressing positive and negative attitudes
toward accommodation requests. Findings show that negative responses from faculty
-------- -----
7
influenced students' decisions in terms of further pursuing support for their disability; in
other words, students were more likely to shy away from disability support if they
experienced negative responses from faculty members when approaching them with
accommodation requests. However, positive responses to accommodation requests from
faculty members led students to further pursue and utilize support services (Hartman-Hall
& Haaga, 2002). In this case, faculty attitudes had a direct impact on students'
motivation to pursue their degrees, as faculty attitudes directly influenced students'
decision to seek out and utilize support services. These findings suggest that faculty
attitudes greatly influence retention rates of students with disabilities.
Dowrick et aI., (2005) found that students with disabilities experienced difficulty
obtaining basic accommodations and supports in various postsecondary settings. Using
focus groups, samples of students with disabilities from varying postsecondary settings
were gathered, including large four-year institutions and two-year community colleges.
These institutions represented different geographic locations within the United States,
including large urban, suburban, and rural areas. A major finding was the students'
perception that disability policy often does not lead to practice; in short, legislation does
not ensure that students will receive the accommodations they rightfully deserve.
Students expressed the need to self-advocate for basic accommodations and described
encounters with faculty members unwilling to accommodate or unknowing of the
disability law. Conversely, students reported faculty mentor relationships were among
the most valuable types of support.
8Inclusive instructional practices. Much of the research has addressed student
perceptions of faculty members' willingness to provide requested accommodations.
However, when asked, students reported that a barrier to learning (and potentially
retention) was more about the instructional practices of faculty members, and less about
their willingness to accommodate (Madaus, Scott & McGuire, 2003a). Madaus, Scott,
and McGuire (2003a, 2003b) conducted a study focused on student perspectives of
effective instructional practices. Students were asked to identify instructional strategies
that had positive effects on them in their college courses, as well as barriers that had
negative effects. Findings show that students feel most successful in courses where clear,
consistent expectations are set from the beginning of the course, learning was treated as a
process, and a variety of instructional strategies were employed by the professor. In
contrast, students reported barriers to learning when professors gave inconsistent
assignments or exams related to material presented in lecture (Madaus et aI., 2003a).
Importantly, these aspects are not specific to accommodation requests; in other words,
clear and consistent expectations, and the use of various instructional practices will
benefit all students, not only students with disabilities.
Further, the study found the main difference between positive and negative
experiences of students with disabilities in college courses was dependent on the attitudes
of individual professors (Madaus et aI., 2003a; 2003b). Professors who promoted
positive learning, were approachable and available, were consistent in communication of
course material and expectations, used engaging instructional strategies (e.g. hands-on
and/or small group activities, scaffolded assignments), made connections to prior
9knowledge or other, more understandable sources, challenged students to think and learn
differently, and positively recognized individuality within students (Madaus et aI.,
2003b). All of these qualities are centered on inclusive instructional practices and are not
necessarily based on accommodations. While some may be linked to accommodations, it
is important to recognize that none of these instructional characteristics are actually
accommodations. Further, all of these qualities could be included as components to
professional development training for faculty.
Evidence shows a primary reason for the low retention rates of postsecondary
students with disabilities is negative attitudes expressed by faculty members (Beilke &
Yssel, 1999; Deshler, Ellis, & Lenz, 1996). These findings indicate that discriminatory
faculty attitudes and assumptions about students with disabilities are a major barrier to
student persistence and retention in four-year colleges and universities. Therefore, it is
imperative to further study faculty attitudes toward and perceptions of students with
disabilities in order to improve retention of students with disabilities at four-year colleges
and universities. In particular, an examination of faculty willingness to provide requested
accommodations and use inclusive instructional practices is warranted.
Faculty attitudes and perceptions. Faculty attitudes and perceptions toward
students with disabilities make a difference because they directly affect the campus
climate and personal experiences of all students. As more students with disabilities enroll
in four-year colleges and universities, it is becoming increasingly urgent to examine these
attitudes in order to design and implement effective interventions that will build
supportive campus climates for all students.
10
Research findings show mixed results of faculty attitudes toward and perceptions
of students with disabilities. Some faculty, staff, and administrators have negative
attitudes and/or lack knowledge about the characteristics of students with disabilities
(Leyser et aI., 1998; Reed et aI., 2003; Vogel et aI., 1999). Parker, Embry, Scott, and
McGuire (2003) conducted a study in which disability services personnel were asked to
describe typical questions from faculty members about teaching students with disabilities.
Responses fell into six major categories of concern: (a) how students with disabilities
could be "otherwise qualified" to participate in higher education; (b) providing effective
instruction to students with disabilities; (c) providing accommodations to students with
disabilities; (d) maintaining academic standards when students with disabilities
participate in higher education, including issues of fairness; (e) characteristics of students
with disabilities; and (f) compliance with legal obligations to assure equal access to
higher education for students with disabilities. Further, study participants represented a
range of faculty knowledge and attitudes about students with disabilities. Some faculty
members had no knowledge of testing or teaching accommodations and were not aware
of existing disability legislation; whereas other faculty members expressed a tendency to
"over-accommodate", asking the disability services personnel what he/she needs to do to
get the student to "pass" (Parker et aI., 2003, p. 7).
Much of the research on faculty attitudes has focused on accommodation requests
from students with disabilities. Vogel et al (1999) found that faculty attitudes can be
negative depending on the type of accommodation request. For example, faculty
responded they were more willing to provide teaching accommodations than exam
11
accommodations, and less willing to provide supplementary materials, such as lecture
outlines or alternative assignments. Faculty members indicated they were most willing to
allow students to record lectures, and to allow extended exam time and to take proctored
exams in another location (such as an Office of Disability Services), but they were least
willing to alter the format of exams (Vogel et aI., 1999). A study by Leyser et ai. (1998)
showed similar results, as well as revealed differences in attitudes and perceptions
according to other variables such as gender, personal experience, faculty rank and
departmental affiliation. Further, and consistent with previous research, Reed et aI.,
(2003) found that faculty showed less support for accommodations and support services
that required additional funding, staffing, and resources.
Other findings indicate that faculty members generally have positive attitudes
toward granting accommodation requests to students with disabilities (Murray, Wren, &
Keys, 2008; Skinner, 2007). In these studies, faculty willingness to accommodate
correlated with several variables, including gender, personal experience, faculty rank, and
departmental affiliation. Murray, Wren and Keys (2008) found that faculty members are
generally willing to make accommodations for students with disabilities, and that faculty
perceptions differ according to faculty rank, gender, and departmental affiliation.
Skilmer (2007) found that faculty willingness to provide accommodations varied
according to departmental affiliation, faculty rank, and type of requested accommodation.
Vogel et ai. (1999) found that many faculty members had limited contact and experience
with students with disabilities in their classes, as well as limited training and knowledge
of disability legislation. Despite these limitations, most faculty members stated they were
12
willing to make accommodations for students with disabilities as requests arose, and
many reported they had previously made such accommodations (Vogel et al. 1999).
Although it is important to consider the mixed results in the research findings -
some studies show negative attitudes and some generally more positive attitudes toward
granting accommodation requests for students with disabilities - it is also necessary to
further examine the differences in terms of departmental affiliation, gender, faculty rank,
and type of requested accommodation. As well, it is important to consider the effect of
professional development training on faculty attitudes and perceptions toward disability,
and confirm whether such training efforts can alter instructional practices to improve the
experiences of students with disabilities in postsecondary settings.
Type ofaccommodation request. Disability type (physical vs. learning disability)
may influence faculty member attitudes toward students requesting accommodations.
Some studies show that faculty members will reasonably accommodate students with
physical disabilities (Beilke & Ysse1, 1999; Blaqua, Rapaport, & Kruse, 1996; Nelson,
Dodd, & Smith, 1990). Oftentimes, these types of accommodations include ensuring
accessibility of buildings, classrooms, and general facilities on campus. However,
students with learning disabilities require very different accommodations; typically, these
students request exam and teaching accommodations, such as extended exam time, a
private exam room, tape-recorded lectures, outlines oflecture notes, and even altered
exam and/or assignment formats. There is evidence that faculty show prejudicial
attitudes towards students with non-visible disabilities and may even, at times, suspect
13
them of "using their disability as a way to gain preferential treatment" (Beilke & Yssel,
1999, p. 365).
More research is needed to further explore differences in disability type and the
effects on faculty willingness to make accommodations. Perhaps, as noted above, faculty
are more willing, in general, to make accommodations for physically disabled students,
and less willing to make accommodations for learning disabled students. In particular, an
instrument is needed that includes distinct items measuring faculty willingness to
accommodate specific types of requests. For example, faculty should be asked if they
would agree to provide lecture materials in large print for visually impaired students, as
well as if they would agree to provide extended exam time to a learning disabled student.
Some faculty members may agree to the former accommodation request but disagree with
the latter, and vice versa. A valid and reliable instrument that makes these distinctions
between accommodations associated with physical or learning disabilities will help us
further understand the nature of faculty attitudes and perceptions toward accommodation
requests.
Faculty training and professional development. Faculty attitudes and
perceptions are influenced by the amount of information they learn about students with
disabilities (Murray, Lombardi, Wren, & Keys, 2009). Typically, faculty may not know
much about the rights of students with disabilities, including the legislation of Section
504 and the Americans with Disabilities Act. Many faculty members rely on the campus
Office ofDisability Services (ODS) to provide verification as to whether a student is
required to receive accommodations due to a disability. Some may argue this is precisely
14
the role of ODS; others may make the case that faculty ought to become familiar with the
law, and at the very least, keep up with recent rulings of federal courts in order to make
the most appropriate accommodation decisions for students in their classrooms (Cope,
2005). In addition, faculty training can serve to foster a solid working relationship
between ODS and faculty, where clear lines of communication are established and
faculty then feel comfortable using ODS as a resource from semester to semester.
Finally, faculty training can increase knowledge of disabilities in general, as some faculty
members have had little to no previous experience with disability in their academic
careers.
Regardless, knowledge and awareness of disability law have become a necessary
training component for university faculty at four-year college campuses due to the rising
enrollment of students with disabilities. Possibly, with more training in instructional
strategies, faculty attitudes toward students with disabilities can become more positive. If
this is the case, training programs play an integral role in enhancing the quality of
instruction of postsecondary faculty and benefitting all students, including students with
disabilities. The outcome could be a more positive postsecondary learning experience
and greater success in postsecondary degree programs for students with disabilities.
Universal design. Some postsecondary institutions have used the concept of
Universal Design (UD) to spearhead faculty training programs. This framework is used
as an instructional backbone to faculty professional development training. UD principles
help encourage faculty members to utilize and embed more inclusive instructional
practices into their course(s). As previously stated, research has shown that students feel
15
most successful in courses where clear, consistent expectations are set from the beginning
of the course, learning was treated as a process, and a variety of instructional strategies
were employed by the professor (Madaus et aI., 2003a). UD principles may help
encourage faculty members to utilize these aspects.
There are seven principles ofUD that were developed by the North Carolina State
University Center for Universal Design, a group of architects, product designers,
engineers, and environmental design researchers (Connell et aI, 1997). The principles
were meant to provide guidance in the usability of environments for a wide range of
learners. The seven principles are listed in Table 1. Although, little empirical evidence
exists to support the effectiveness of Universal Design (UD), as it is a rather new concept
in the field of postsecondary faculty and support for students with disabilities, the
framework has the potential to influence instructors to use inclusive teaching methods
that reach a broad range of learners. There is an underlying connection between UD and
disability accommodations; both encourage accessibility to more diverse groups of
people. While UD is not as specific as accommodations, it is likely that if the UD
principles are followed, the need for accommodations may decrease (Ketterlin-Geller &
Johnstone,2007). In other words, by encouraging faculty to embed the UD principles
within their course framework, they may experience a decrease of specific
accommodations from students with disabilities. In essence, the accommodations would
come naturally along with the inclusive learning environment promoted by UD.
16
Table 1
Seven Principles ofUniversal Design
Principle
Equitable use
Flexibility in use
Simple and intuitive use
Perceptible information
Tolerance for error
Low physical effort
Size and space for approach
and use
Definition
The design is useful and marketable to people with diverse
abilities.
The design accommodates a wide range of individual
preferences and abilities.
Use of the design is easy to understand, regardless of the
user's experience, knowledge, language skills, or current
concentration level.
The design communicates necessary information
effectively to the user, regardless of ambient conditions or
the user's sensory abilities.
The design minimizes hazards and the adverse
consequences of accidental or unintended actions.
The design can be used efficiently, comfortably, and with
a minimum of fatigue.
Appropriate size and space is provided for approach,
reach, manipulation, and use regardless of the user's body
size, posture, or mobility.
Summary of the problem. The number of students with disabilities entering
postsecondary institutions continues to increase. As this population grows, it is
imperative that retention rates improve so that students with disabilities receive a quality
higher education and have equal opportunities entering the work force as adults.
Although many factors could be identified as barriers to the success of students
with disabilities at four-year colleges and universities, one known barrier is negative
faculty attitudes and perceptions of disability, which is often specifically related to
17
accommodation requests (Leyser et aI., 1998; Reed et aI., 2003; Vogel et aI., 1999).
Some research exists regarding the prevalence of negative and positive attitudes toward
students with either physical or learning disabilities (Leyser et aI., 1998; Reed et aI.,
2003; Vogel et aI., 1999), and some research exists about the differences in faculty
attitudes and perceptions of disability type and willingness to provide requested
accommodations (U. S. Department of Health and Human Services, 2003; U.S.
Department of Labor, 2007; Wagner et aI., 2005). There is some correlational evidence
linking faculty attitudes to other variables such as gender, faculty rank:, and departmental
affiliation (Murray et aI., 2008; Skinner, 2007).
Finally, there is evidence that faculty members who receive training in disability
awareness are more likely to demonstrate positive attitudes and perceptions toward
students with disabilities (Murray et aI., 2009). Possibly, training faculty members in
disability awareness and inclusive instructional principles, promoted by such frameworks
as Universal Design, may help to create more positive attitudes and perceptions of
students with disabilities as well as enhance instructional practices that reach a broad
range of student learners. Training should also include a general overview of disabilities
and specifics about the role of ODS, including strategies for faculty to foster
collaborative working relationships with ODS staff. If faculty training opportunities are
proven to be effective in influencing positive attitudes then appropriate training
interventions should be designed and implemented.
Despite increased understanding about university faculty attitudes and perceptions
towards students with disabilities, prior research suffers from several measurement and
18
methodological limitations. For example, many researchers have not used reliable and
valid measures of faculty attitudes and perceptions when assessing these perceptions and
most analyses focus on item level results rather than reliable latent constructs. Thus,
there is a need to develop and field test reliable and valid measures of faculty attitudes
and perceptions towards university students with disabilities.
Measuring Faculty Attitudes and Perceptions toward Disability
Among the research studies previously described, many different methods were
employed to measure faculty attitudes and perceptions. This methodological variation
indicates a lack of measurement consistency across four-year college and university
settings. In order to further investigate the role of faculty attitudes on the enrollment and
retention rates of students with disabilities at four-year colleges and universities, a
reliable and valid instrument should be developed. The instrument must measure the
identified constructs, based on the literature, that capture faculty attitudes and
perceptions. Also, this instrument should measure the effects of training on faculty
attitudes so that group comparisons can be made between faculty who have and have not
received training. Once developed and validated, the survey instrument can be used
across four-year college and university settings to give administrators a sense of where
and how to target interventions for faculty professional development in the area of
disability awareness.
Some research studies have examined existing measures of attitudes toward
students with disabilities. Not all studies addressed validity and reliability of the
specified measure in the study (e.g., Chubon, 1992; Rao, 2004), such as the Attitudes
19
toward Disabled Persons Scale (ATDP) by Yuker, Block, and Campbell (1960).
Developed in the late 1950's, the ATDP is a measure of attitudes toward persons with
physical disabilities. Although the measure is over 40 years old, recent research studies
have used the instrument and found evidence of adequate reliability, where Cronbach's a
is equal to or greater than .80 (Rao, 2002).
An instrument developed by Askamit, Morris, and Luenberger (1987) included
subscale items that measured comfort levels when working with students with learning
disabilities, reactions to curricular and instructional accommodations, perception of
success rates of students with disabilities, familiarity with disability legislation,
characteristics of leaming disabilities, support services available on campus, and ability
to respond to the needs of students with disabilities (as cited in Rao, 2002). A
Cronbach's a of .82 was reported for the attitude subscale. Although this measure
provides evidence of reliability, it is limited because it only addresses learning
disabilities. Ideally, a measure would include items that measure attitudes toward
learning and physical disabilities so that comparisons can be made as to whether faculty
members treat various accommodation requests in the same manner.
Rao (2002) conducted a literature review of studies on faculty attitudes of
students with disabilities in higher education, in which attitudinal measures were
included. Common variables that were included across studies were (a) gender, (b) age,
(c) experience, (d) rank, (e) departmental affiliation, (f) knowledge of disability laws, and
(g) disability type (Rao, 2002). Rao found that gender had an effect on willingness to
accommodate in five research studies included in the literature review, and no effect on
20
willingness in four studies. Only one study (Fonosch & Schwab, 1981) showed faculty
rank as a significant factor in faculty willingness to accommodate; other studies showed
no difference in faculty rank (e.g. assistant, associate, or full professor, and instructor).
Departmental affiliation, however, did make a difference in faculty attitudes in five of the
studies, which included findings that faculty from "soft sciences" (e.g. sociology,
psychology) were more likely to have positive attitudes than faculty from the "hard
sciences" (e.g. math, biology), and faculty from the Education Department tended to
show more willingness to accommodate than other departments (Rao, 2002).
Murray, Wren and Keys (2008) developed the Project PLUS survey to measure
faculty attitudes and perceptions of students with learning disabilities. The 38-item
instrument was given to faculty at a private Midwestern university located in an urban
setting. An exploratory factor analysis yielded 12 reliable factors, which included (a)
willingness to provide major accommodations, (b) willingness to provide exam
accommodations, (c) fairness and sensitivity, (d) knowledge oflearning disabilities, (e)
willingness to personally invest, (f) willingness to provide teaching accommodations, (g)
resource constraints, (h) performance expectations, (i) disclosure and believability, CD
inviting disclosure, (k) knowledge to make accommodations, and (1) providing
accommodations. Although the 12 factors addressed a range of constructs, the instrument
is limited to measuring faculty attitudes toward learning disabilities; willingness to
provide accessibility-related accommodations is not included as a construct in the
measure.
21
In general, these measures have some commonalities. Even though some
measures focused on physical disabilities and some on learning disabilities, the measures
had constructs in common, including (a) perceptions and knowledge of disability (visible
and non-visible), (b) willingness to invest time supporting students with disabilities, (c)
fairness and sensitivity, (d) performance expectations of students with disabilities, (e)
knowledge of disability law, (f) willingness to provide teaching, exam, and accessibility
accommodations, and (g) knowledge of support services and campus resources targeted
toward students with disabilities. Further development of an instrument that measures
faculty attitudes and perceptions should include items that capture these general
constructs and address different disability types.
More research is needed to further determine and develop adequate, reliable, and
valid instruments. Since evidence exists that students with disabilities face additional
challenges and low retention and completion rates at the postsecondary level, it is crucial
to further explore the possible barriers to success faced by this population. Since faculty
attitudes affect students with disabilities in their pursuit for a college degree, the ability to
measure these attitudes will help colleges and universities pinpoint further interventions
to solve the problem, including directed training and professional development efforts.
Also, once training is implemented, measures ought to capture the effects of training on
faculty attitudes and perceptions toward students with disabilities in order to evaluate
interventions.
22
Establishing validity. An important step of developing an adequate survey
instrument is establishing construct validity, which is the degree to which the instrument
measures what it purports to measure (Haynes, Richard, & Kubany, 1995; Kane, 2001;
Messick, 1980). Content validity, a facet of construct validity, centers on developing and
refining the constructs of the instrument by reviewing the measure blueprint,
specifications, and the measure itself, in order to provide evidence based on test content
as well as contribute to improved clarity of the measure (AERA, APA, & NCME, 2009;
Haynes, Richard, & Kubany, 1995).
In this case, an adequate instrument that measures faculty attitudes should include
items meant to target and measure specific underlying theoretical constructs. Specifically,
the theoretical constructs include (a) perceptions and knowledge of disability (visible and
non-visible), (b) willingness to invest time supporting students with disabilities, (c)
fairness and sensitivity, (d) performance expectations of students with disabilities, (e)
knowledge of disability law, (f) willingness to provide teaching, exam, and accessibility
accommodations, and (g) knowledge of support services and campus resources targeted
toward students with disabilities. An additional factor could be (h) willingness to utilize
inclusive instructional practices, as Madaus et al. (2003a) found that students with
disabilities cited the most success with these types of courses. These eight theoretical
constructs must be observed and tested in order to determine ifthey adequately represent
faculty attitudes. Validating these eight theoretical constructs through analysis of an
assessment instrument is essentially the process of building construct validity. Most
importantly, validation of the interpretation of the test results is crucial; not validation of
23
a test score itself, but the interpretation and utilization of that test score, and the decisions
that will be made based on that test score that will affect future outcomes (Kane, 2001).
In this case, the results will help inform postsecondary institutions of faculty attitudes and
perceptions, and appropriate targeted interventions, such as professional development
training in disability awareness, will be the outcome.
In developing an assessment instrument, Clark and Watson (1995) provide
recommendations for procedures to follow in establishing construct validity. These steps
include (a) the identification of a theoretical model, (b) building substantive validity,
which includes the development of an initial item pool, and (c) building structural
validity, which includes item selection and psychometric evaluation (Clark & Watson,
1995). As previously mentioned, theoretical constructs have been identified after review
of the literature, and item construction is based on eight hypothesized factors. A
complete validity study of a survey instrument that measures faculty attitudes and
perceptions will include the creation of an initial item pool, item selection, and
psychometric evaluation using methods such as factor analysis.
Factor analysis. In order to establish content and structural validity, the items
that are meant to measure these constructs may be psychometrically evaluated with the
methods exploratory factor analysis (EFA) and confirmatory factor analysis (CFA).
Applying such methods will confirm or disconfirm the instrument constructs, and inform
research as to whether they need to be refined or revised (Clark & Watson, 1995). Most
importantly, this process will help to ensure the instrument measures what it purports to
measure, the basic definition of construct validity (Kane, 2001; Messick, 1980). Both
24
EFA and CFA are more robust that other methods (MANOVA, for example) because
they allow the researcher to detemline goodness of model fit using theoretical constructs
measured by a group of items, rather than one single item. Groups of items (called
factors) are written to capture one holistic construct and then tested for validity in the
factor analysis process.
EFA helps detemline the nature oflatent variables that explain the variation and
covariation in a set of measured variables by revealing the sources of common variation
underlying the measured variables within the data (Preacher & MacCullum, 2003). The
results of the EFA will determine the appropriate item groupings and develop the
unknown constructs of the instrument that gauge attitudinal changes in faculty toward
students with disabilities. It is imperative to adequately conduct the EFA on the survey
instrument otherwise erroneous conclusions may be drawn regarding faculty attitudes
about students with disabilities. Factor analysis can only be a useful tool if used correctly
and appropriately; unfortunately some researchers end up using results foolishly by
misinterpreting item groupings and misrepresenting constructs (Clark & Watson, 1995).
In order to ensure adequate interpretation of item groupings and constructs, researchers
must carefully consider (a) extraction, (b) the number of factors to retain, and (c) rotation
(Costello & Osborne, 2005).
Confirmatory factor analysis (CFA) is a method where the factor structure
resulting from an EFA can be tested and confirmed. Researchers use CFA instead of
EFA when there is a strong rationale or hypothesis as to why the items are grouped or
correlated. Some studies will employ both EFA and CFA methods by splitting the
25
sample size in half and then assigning the first half to EFA and the second half to CFA
(Johnson, Stevens & Zvoch, 2007: Johnson & Stevens, 2001). The factors extracted in
the EFA can be confirmed in the CFA for the remaining portion of the sample. The CFA
models are then estimated using structural equation modeling (SEM) methods. In
measurement studies, SEM is more ideal than standard statistical procedures (such as
ANOVA) because SEM involves a priori hypotheses and offers the convenience of
differentiating between observed and latent variables (Kline, 1998). Because research
exists on some measures of faculty attitudes and perceptions of students with disabilities,
and preliminary constructs have been identified based on these pre-existing measures,
this is a case where hypotheses have been made. It is appropriate, in this case, to employ
both EFA and CFA because the constructs are still exploratory in nature and the use of
both methodologies will offer more robust results.
Research questions. The primary purpose of this investigation was to develop
and field test a survey instrument intended to measure faculty attitudes and perceptions
toward students with disabilities at four-year colleges and universities. A secondary
purpose was to examine how demographic characteristics of the sample and prior training
experiences influenced faculty attitudes and perceptions. The research questions were as
follows:
(a) Is there evidence of reliability and validity for the ExCEL survey?
(b) Are there group differences by gender, faculty rank, and departmental
affiliation?
(c) Are there group differences by prior disability awareness training?
26
CHAPTER II
METHODS
This study explored the validity and reliability of the Expanding Cultural
Awareness of Exceptional Learners (ExCEL) survey instrument in order to determine ifit
can be used across four-year college and university settings. Along with evaluation of
psychometric properties, group differences were examined according to gender, teaching
status, and college affiliation. As well, group differences were examined to determine if
training has an effect on faculty attitudes and perceptions toward disability.
Instrument
A pre-existing instrument, the PLUS survey, (Murray et aI., 2008) was adapted
and revised into a new instrument, the Expanding Cultural Awareness ofExceptional
Learners (ExCEL) survey. It is important to note that the PLUS survey was used to
measure faculty attitudes and perceptions toward students with learning disabilities,
whereas the ExCEL survey included items measuring faculty attitudes and perceptions
toward all students with disabilities. Therefore, the initial item pool for the ExCEL
survey included extracted items from the PLUS survey and newly drafted items based on
the literature. The ExCEL survey is included in Appendix A.
Important decisions were made during the development of the ExCEL survey.
The original twelve reliable factors from the pre-existing PLUS survey were retained, but
27
some modifications were made. These changes were agreed upon after a consultation
meeting with the Director of Disability Services. In an effort to create a more
parsimonious instrument, three factors related to accommodations were collapsed into
one "Willingness to Make Major Accommodations", "Willingness to Make Exam
Accommodations", and "Willingness to Make Teaching Accommodations". One factor
was added to the hypothesized list: Willingness to Create More Inclusive Learning
Environments. Nine items were written to capture this construct. This decision was
made in particular because Disability Services had been working to educate faculty
members on the principles of Universal Design, and many of these principles include the
concept of delivering instruction in an inclusive fashion. As such, faculty members have
been encouraged to embed more inclusive instructional strategies into their course
curricula. They are not required to do so, but Disability Services would like to have a
sense of how many faculty members are willing to make these curricular changes.
The instrument included 57 selected response questions with ordered response
categories meant to measure level of agreement or opinion (Dillman, 2007). This format
was chosen because the ExCEL survey is meant to uncover the attitudes and beliefs of
the faculty members. By selecting a vague quantifier to encompass their opinion of a
topic, participants may quickly respond to opinion questions on topics they mayor may
not have previously considered (Patton, 1982). There were six response options -
strongly agree, agree, somewhat agree, somewhat disagree, disagree, and strongly
disagree.
28
Participants
Participants included teaching faculty at a medium-sized public research
university located in the Pacific Northwest. As of 2007, there were 1,714 faculty
members and 1,267 graduate assistants at this university, a total of2,98l teaching faculty.
Eighty-two percent of the faculty members were white, seven percent were Asian/Pacific
Islander, three percent were Hispanic, one percent was African American, one percent
was Native American, and one percent was Multi-ethnic. Five percent declined to report
racial identity. Fifty-four percent of faculty members were male and forty-six percent
were female. The average faculty-to-student ratio for the university was 1: 18.
At this university, the number of students with disabilities has grown
approximately 20% over the past five years. As of2007, there were 20,376 students
emolled, including 763 students with disabilities who comprise 4% of the student
population. Of the students with disabilities population, 63% are diagnosed with either a
learning disability or Attention Deficit Hyperactivity Disorder. The remaining 37% are
diagnosed with another disability type, such as mobility, hearing, visual, speech
impairments, psychological or health disability, brain injury, or seizure.
Due to budgetary constraints with the survey incentives, the decision was made to
recruit only those faculty members employed at .5 full-time employment (FTE) or
greater. This decision was made with the assumption that these faculty members would
have the greatest impact on the students they teach. After selecting only those faculty
members employed at .5 FTE or greater, the survey was sent to 1,084 faculty members.
As a result, no graduate assistants were recruited for the study.
29
I met individually with each Dean from the eight separate colleges. After giving
an overview of the project, I asked for their help in providing me with access to the email
listserv used to send information to faculty members employed .5 FTE or greater within
their respective college. After meeting with the Dean of the Law School and the Director
of Disability Services, it was determined the Law School faculty would not receive the
surveyor participate in the study. This decision was made because different procedures
in providing accommodations to students with disabilities are followed at the Law
School, and essentially the Disability Services office handles their requests in a much
different manner. For example, Law School students do not submit any accommodation
requests directly to faculty. Instead, Disability Services staff work with administrative
staff at the Law School to accommodate students with disabilities. Because of these
differences, many of the items on the survey would not be relevant to Law School
faculty. After all meetings with the Deans occurred, the finalized participant population
included faculty from all of the colleges (with the exception of the Law School) who
were employed at .5 FTE or greater. This population included 1,084 potential
participants.
Procedures
Faculty members received an email including an introductory recruitment letter
that described the research project and a link to the ExCEL survey (see Appendix B),
which was presented in an online format through Survey Monkey. They were asked to
complete the survey on a voluntary basis. All participants were offered a $3 coupon to a
popular local cafe and bakery, regardless ofwhether they completed the surveyor not.
30
Participants completed an online consent form prior to participating in the actual survey
(see Appendix C). Ifparticipants selected "no" to the consent form, they were not
allowed to advance to the beginning of the online survey. Participants were sent three
follow-up emai1s encouraging them to take the survey after the initial email invitation
was sent (see Appendices D, E, and F). These emai1s were sent two, four, and six weeks
after the initial email invitation was sent, respectively. In order to track faculty responses
over time, a series of security questions were added at the beginning of the survey (see
Appendix G).
Sample and Response Rate
At the conclusion of the data collection period, the final sample consisted of 295
participants. The overall response rate was 27%. Although far from ideal, it is consistent
with response rates of previous studies of similar nature involving university faculty and
disability awareness (Bourke et al., 2000; Murray et al., 2008; Murray et al., 2009; Vogel
et al., 1999, Vogel e1 al., 2008). The highest response rate was from the Honors College
(71 %) although this particular college had the smallest population. The lowest response
rate was from Journalism (18%). Table 2 lists the response rate by college membership.
Figures 1 and 2 illustrate the differences between the population and representative
sample.
31
Table 2
Sample Characteristics and Response Rates ofFaculty by College Membership
All Faculty Sample Response Rate
College N % N %
Arts and Sciences 490 45% 137 47% 28%
Architecture and Allied Arts 148 14% 32 11% 22%
Business 81 7% 32 11% 40%
Journalism 60 6% 11 4% 18%
Honors 7 < 1% 5 1% 71%
Education 243 22% 48 17% 20%
Music and Dance 55 5% 25 9% 45%
Total Participants
III Arts dnd Sciences
mAAA
Business
till Journdlism
Honor:.
Education
Musicand Ddnce
Figure 1. Representation of all possible faculty participants
32
Sample
mil Arts and Sciences
flJAAA
fIJ Journalism
Iionors
EdLJl:ation
Mw;ic and Dance
Figure 2. Representation of faculty sample included in the study
After subjective inspection of pie charts representing the population and sample, I
determined, for the most part, the sample was representative of the faculty population,
with some overrepresentation of Business, Honors College, and Music faculty.
In addition to college membership, participants were asked to report gender, age,
teaching status, years teaching, and type of courses taught. For gender, 44% ofthe
faculty were male and 56% were female. The mean reported age was 48.3 (SD = 10.2).
For teaching status, 64% of the faculty reported they were tenure-line, indicating they
were tenured or on a tenure-track, whereas the remaining 36% reported they were
nontenure, indicating they were nontenure track, visiting, or post-retire. The mean
reported years teaching at the postsecondary level was 15.86 (SD = 10.1). When asked
which type of courses they primarily taught, 55% of participants indicated a major-
specific course, 25% indicated a graduate-level course, 12% indicated a general
33
education course, 6% indicated an elective course, and less than 1% indicated a lab or
discussion section.
Additionally, participants were asked a series of questions about prior experience
in disability awareness and teaching students with disabilities. When asked how many
students with disabilities they had taught in the past five years, 37% of participants
reported 1-5 students, 22% reported 6-10 students, 12% reported 10-20, 12% reported
more than 20, 13% reported "don't know or not sure", and 4% reported "none".
Participants were asked if they had a disability, and 9% responded "yes"; meanwhile 49%
indicated a family member, friend, or other personal contact had a disability, 72%
indicated they had taught at least one student with a disability at some point in their
teaching career, and 14% indicated they had no personal experiences with disability.
Finally, participants were asked if they had ever received training related to
disability or working with college students with disabilities, where 29% reported "Yes"
and 71 % reported "No". Of the 29% who had received some type of disability-focused
training (n = 86),64% reported they attended a workshop, 27% reported they took one or
more courses, 60% reported they read books or articles, 38% reported they visited
website(s), and 27% reported "other" (please note: participants were asked to check all
that apply for type of training received).
Data Analyses
Examination of reliability and validity addressed research question one.
Reliability was examined using Cronbach's u, a measure of internal consistency, on the
entire instmment and within subscales. Criterion of.70 or higher was considered
34
adequate reliability, and .80 or higher was considered preferable (Nunnally, 1975).
Validity was examined through the psychometric evaluation of the instrument using
exploratory and confirmatory factor analysis.
In exploratory factor analysis (EFA), there are three important decisions to
consider: (a) the extraction method (b) the number of factors to retain, and (c) the rotation
method to be utilized (Costello & Osborne, 2005; O'Connor, 2000; Preacher &
MacCullum, 2003). In the first decision, Principal Axis Factoring (PAF) was selected as
the extraction method. In the second decision, Kaiser's rule and scree plots were used to
determine the number of factors to retain. Kaiser's rule states that only eigenvalues
greater than 1.0 shall be retained, as these components tend to account for more variance
of a single measured variable. As well, parallel analysis and Velicer's minimum average
partial (MAP) test were utilized; these procedures have been validated, unlike Kaiser's
rule and scree plots (O'Connor, 2000), and therefore using a combination of these
procedures to determine the number of factors to retain is optimal. The third decision
involved choosing between orthogonal and oblique rotation methods. Essentially,
orthogonal rotation requires the factors to be uncorrelated, whereas oblique methods do
not employ such a restriction. In this study, it was most appropriate to employ oblique
rotation methods since this method allows the estimation of factor cOlTelations, and I
suspected the constructs would be interrelated.
In carrying out the psychometric evaluation of the instrument, there were two
possible plans for data analyses presented in my proposed dissertation, which I referred to
as Plan A and Plan B. Plan A was the ideal data analyses plan, and included conducting a
35
crossvalidation study with exploratory and confirmatory factor analysis. The sample was
to be randomly split so that half was subj ect to EFA and the other half subj ect to the
confirmatory factor analysis (CFA). This process would test the validity of the instrument
and therefore address the second part of research question one. Research questions two
and three would be addressed through invariance testing, where the factor structure of the
instrument would be compared according to the stated predictor variables (gender,
teaching status, college membership, previous training). Plan A was to be utilized ifthe
participant response rate of the survey was reasonable and a minimum of a 2: 1 case-to-
variable ratio was attained (Costello & Osborne, 2005; Fabrigar, Wegener, MacCallum,
& Strahan, 1999; Stevens, 2002).
Like Plan A, Plan B included exploratory factor analysis (EFA). However, Plan B
was to be utilized if the desired 2: 1 case-to-variable ratio was not attained due to a low
response rate, and the sample size was not large enough to conduct a CFA (Kline, 1998).
Invariance testing is not possible without conducting a CFA, and therefore Plan B was to
address research questions two and three by conducting Multivariate Analysis of
Variance (MANOVA) to determine the group differences in responses on the factor
scores (taken from the EFA results) according to the stated predictor variables (gender,
teaching status, college membership, previous training). Essentially, since Plan B
addressed the situation where sample sizes would be too small to conduct CFA,
MANGVA was substituted for the group comparisons, and would address research
questions two and three.
36
The sample size was large enough to conduct both an EFA and CFA; however, it
was not large enough to conduct invariance testing. A minimum sample size of 200 is
required per group to conduct invariance testing (Kline, 1998). Therefore, a combination
of Plan A and B was followed, where the EFA and CFA were conducted and one-way
between-subjects MANOVAs were conducted to examine group differences. In the
MANOVAs, factor scores were the outcome variables and the categorical predictor
variables were gender, faculty rank, college membership, and training status (whether or
not the participant received previous training on disability awareness). In total, three
MANOVAs were conducted. The a priori criterion for declaring statistical significance
was .05.
As previously stated, reliability was examined using Cronbach's a on the entire
instrument and within subscales. Subscales were determined based on the EFA results,
and the subscale scores were used as the outcome variables for the MANOVAs. The
MANOVAs were conducted using the CFA sample (n = 200).
37
CHAPTER III
RESULTS
The purpose of this study was to test the initial reliability and validity of an
instrument intended to measure the attitudes and perceptions of university faculty
members toward disability. The measure was a 57-item survey that was designed with
eight subscales. In this study, I sought to answer the following research questions:
Research Question 1: Is there evidence of reliability and validity for the ExCEL
survey?
Research Question 2: Are there group differences by gender, faculty rank, and
departmental affiliation?
Research Question 3: Are there group differences according to the amount of
disability awareness training received?
Analyses
As mentioned in the previous chapter, there were two potential data analysis plans
outlined (Plan A and Plan B), and the final decision between them was dependent on
sample size. The sample size was large enough for Plan A to be carried out; however, the
sample size was not large enough to conduct invariance testing, which would have
addressed research questions two and three in Plan A. Due to the insufficient sample
size, multivariate analysis of variance (MANOVA) was utilized to examine group
-------------
38
differences in order to answer research questions two and three. It is important to note
that while MANOVA is a robust methodology in testing for group differences, it is
particularly different from invariance testing because means are compared in MANOVA,
whereas with invariance testing the factor structure of the instrument is compared
according to group membership. Therefore, invariance testing provides more information
on the ways in which structure differs by group, such as correlations among factors,
coefficients, and residuals)
Two types of software were used for analyses, SPSS 15.0 (SPSS, Inc., 2006) and
AMOS 7.0 (Arbuckle, 2006). Prior to conducting analyses, data were screened for model
assumptions of multivariate normality and homoscedasticity. Model assumptions were
met, and there were some items with non-nonnal distributions. Because of this, principal
axis factoring (PAF) was chosen as the factor extraction method over maximum
likelihood (ML) (Costello & Osborne, 2005). Outliers were present, and mean
differences were examined before and after removing the outliers from the data. Some
outliers were more than 3 standard deviations from the mean, and were subsequently
flagged as potential influential cases (Stevens, 2002). Approximately 8% of the data
were missing. Missing data were treated with imputation using the estimation
maximization (EM) algorithm (Dempster, Laird, & Rubin, 1977).
Psychometric properties of the instrument were examined in the exploratory and
confirmatory factor analysis results. Because this was a crossvalidation study, the initial
total sample (n = 295) was randomly split so that approximately one third of the sample
was used for the EFA (n = 95), and approximately two thirds were used for the CFA (n =
39
200). The choice to split the sample accordingly was based on a minimum sample size
recommendation of 200 for confirmatory factor analysis (Kline, 1998).
Exploratory factor analysis. In total, three EFAs were conducted to determine
the best factor structure to use in the CFA model. For each EFA, PAF was used with
oblique rotation and the same decision rules were applied concerning item deletion.
According to these decision rules, an item was removed if: (a) it cross-loaded on more
than one factor at .32 or greater, (b) it failed to load onto any factors, (c) it had a
communality less than .20, or (d) the item loaded onto a factor, but after close
examination, it was determined the item did not group well conceptually with other items
in the factor, and therefore the construct was not theoretically plausible (Costello &
Osborne, 2005; Tabachnick & Fidell, 2001).
The initial EFA included all items in the original survey (n = 57, see Appendix
A). After applying the aforementioned decision rules, eighteen items were removed and
another EFA was conducted with the remaining items (n = 39). After applying the
decision rules again, twelve items were removed and another EFA was conducted with
the remaining items (n = 27). The 27-item version of the survey instrument was the final
EFA conducted. More items were removed prior to conducting the CFA based on the
same decision rules and the final model used for the CFA included a reduced number of
items (n = 19).
The means and standards deviations of each item are included in Appendix H.
The correlation matrix for the items is shown in Appendix 1. Item intercorre1ations were
low to moderate. The item communalities were moderate to strong, with values ranging
40
from the highest item "I am confident in my understanding of the Americans with
Disabilities Act" (.899) and the lowest item "I routinely build in some flexibility in my
evaluation methods, such as dropping the lowest quiz score or providing different exam
formats (essay, multiple choice)" (.120). This item was the least correlated with the other
items, and was so low it was not retained. There was one other item that fell below the
0.20 criterion, and therefore was not retained ("I refer students to academic support
services on campus, such as Academic Learning Services (ALS), when they need help")
(.196). The item communalities are shown in Table 3. Immediately following, factor
eigenvalues are summarized in Table 4.
Table 3
Communalities ofExCEL Survey Items
Item
I am confident in my understanding of section 504 of the
Rehabilitation Act of 1973
I am confident in my understanding of the legal definition of disability
I am confident in my understanding of the Americans with Disabilities
Act (1990)
I am willing to provide copies of my lecture notes or outlines to
students with documented disabilities
I am willing to provide copies of my overheads and/or PowerPoint
presentations to students with documented disabilities
Providing accommodations to students with disabilities is a way to
ensure equal opportunity and access to learning in higher education
settings
I am willing to spend extra time (i.e., in addition to typical office
hours) meeting with students with documented disabilities to clarify
and/or review course related content or prepare for an upcoming exam
Extraction
0.684
0.797
0.899
0.606
0.706
0.384
0.776
Item Extraction
I am willing to allow a student with a documented disability to 0.858
complete extra credit assignments for academic success even when this
option is not stated on the course syllabus
I am willing to allow any student to complete extra credit assignments 0.561
in my course(s)
Providing teaching accommodations to students with documented 0.592
disabilities is unfair to students without disabilities
I believe that students with disabilities use the disability as an excuse 0.692
when they are not doing well in my class
I am willing to arrange extended time on exams for students who have 0.861
documented disabilities
I am willing to spend extra time (i.e., in addition to normal office 0.774
hours) helping any student prepare for an exam or review course
material
I feel very comfortable meeting with students to discuss their 0.724
disability-related accommodation needs
I am willing to allow students with documented disabilities to take 0.434
proctored exams in a supervised location outside of the normal exam
location
I feel very comfortable discussing with students barriers they may 0.454
experience given the design or evaluation methods used in my
course(s)
I routinely build in some flexibility in my evaluation methods, such as 0.120*
dropping the lowest quiz score or providing different exam formats
(essay, multiple choice, papers) for all students
Students with documented disabilities who request support from 0.610
Disability Services receive adequate services from that office
I receive adequate support from Disability Services to make 0.786
appropriate accommodations for students with documented disabilities
Making adequate accommodations for students with documented 0.601
disabilities in my courses is unrealistic given time constraints and other
job demands
I include a statement in my syllabus inviting students with disabilities 0.541
41
Item
to discuss their needs with me
I make a statement in class inviting students with disabilities to discuss
their needs with me
I would like to modify curricular aspects of my course(s) in order to
create a more inclusive learning environment for all students
I prefer to use a variety of instructional formats in my class, including
small group and hands on activities
I believe it is my responsibility as an instructor to provide consistent
information and expectations on all assignments and exams
I refer students to academic support services on campus, such as
Academic Learning Services (ALS), when they need extra help
Students with disabilities should be able to perform just as well as
students without disabilities in my course(s)
*communality < .20 and item removed from CFA
Extraction
0.456
0.588
0.375
0.386
0.196*
0.405
42
43
Table 4
Total Variance Explained of Unrotated Eigenvalues and Oblique Rotation Eigenvalues
Unrotated Oblique
Rotation
Eigenvalue Percent of Cumulative Eigenvalue
Factor vanance percent
Knowledge of Disability Law 5.722 21.2% 21.2% 3.166
Fairness 2.594 9.6% 30.8% 3.449
Invitation to Disclose 2.227 8.2% 39.0% 2.975
Willingness to Invest Time 1.848 6.8% 45.9% 2.985
Accessibility of Course 1.658 6.1% 52.0% 2.361
Materials
Willingness to Adjust Course 1.470 5.4% 57.4% 1.723
Assignments
Satisfaction with Campus 1.398 5.2% 62.6% 1.854
Resources
Willingness to Provide Testing 1.142 4.2% 66.9% 2.511
Accommodations
Eight factors emerged from the final EFA. The decision to retain eight factors
was based on the more conventional methods of the Kaiser-Guttman rule (eigenvalues>
1) and examination of scree plots (Preacher & MacCallum, 2003), as well as Velicer's
MAP test and Parallel Analysis (O'Connor, 2000), all of which showed slightly different
results. According to Kaiser's rule and examination of the scree plot, eight factors
emerged. However, the results of Velicer' s MAP test revealed seven factors to retain,
and the results of the Parallel Analysis test revealed nine factors. Since Velicer's MAP
44
test tends to underestimate the number of factors, and Parallel Analysis tends to
overestimate (O'Connor, 2000), I made the decision to retain eight factors based on the
combination of results. The factors were labeled: (a) "Knowledge of Disability Law", (b)
"Fairness", (c) "Invitation to Disclose", (d) "Willingness to Invest Time", (e)
"Accessibility of Course Materials", (f) "Willingness to Adjust Course Assignments", (g)
"Satisfaction with Campus Resources", and (h) "Willingness to Provide Testing
Accommodations". With these eight factors, approximately 67% of the total variance of
the original items was accounted for. Prior to rotation, eigenvalues were unequally
distributed, but after applying rotation, smoothing occurred from uillotated to rotated
solutions and eigenvalues were redistributed.
Due to moderate intercorrelations between items, pattern coefficients seemed
most appropriate to report and interpret. Results of the pattern matrix of coefficients are
shown in Table 5.
Table 5
Pattern Coefficients Resulting/rom EFA with PAF and Oblique Rotation
Item
Q I: understanding of section
504
Q2: understanding of
disability
Knowled Fairness Invitation Willing to Accessib- Willing to Satisfaction Willing to
ge of to Invest ilityof Adjust with Provide
Disability Disclose Time Course Course Campus Testing
Law Materials Assignments Resources Accom-
modations
0.795
0.859
Q3: understanding of ADA
Q13: accommodations ...
ensure equal opportunity and
access
Q22: ...providing
accommodations ...unfair to
students without disabilities
Q24: students with
disabilities ... use the
disability as an excuse
Q41: Making adequate
accommodations .,. is
unrealistic given time
constraints .
Q46: statement in my
syllabus inviting students
0.942
0.425
-0.835
-0.868
-0.488
0.726
..j::>.
Vl
Item
Q47: statement in class
inviting students
Q14: willing to spend extra
time ... meeting with
students with disabilities
Q27: willing to spend extra
time ... helping any student
QlO: willing to provide
copies of my lecture notes
Q12: willing to provide ...
overheads and/or
PowerPoint
Q15: willing to allow
students with disabilities
complete extra credit
assignments
Q16: ... allow any student to
complete extra credit
assignments
Q38: Students with
disabilities ... receive
adequate services ...
Disability Services
Knowled
ge of
Disability
Law
Fairness Invitation
to
Disclose
0.584
Willing to
Invest
Time
0.837
0.867
Accessib-
ilityof
Course
Materials
0.790
0.790
Willing to
Adjust
Course
Assignments
0.925
0.690
Satisfaction
with
Campus
Resources
0.790
Willing to
Provide
Testing
Accom-
modations
~
0\
Item
Q39: I receive adequate
support from Disability
Services
Q25: willing to arrange
extended time on exams
Q29: willing to allow ...
proctored exams in a
supervised location
Knowled
ge of
Disability
Law
Fairness Invitation
to
Disclose
Willing to
Invest
Time
Accessib-
ilityof
Course
Materials
Willing to
Adjust
Course
Assignments
Satisfaction
with
Campus
Resources
0.886
Willing to
Provide
Testing
Accom-
modations
0.941
0.562
Note: Coefficients < .30 were removed from the table for ease of interpretation
.j::>.
-..l
48
Based on the results of the pattern matrix, I carefully examined the conceptual
grouping of items within factors. From this process, I determined eight factor labels that
seemed to express the most logical communality of the items as theoretical constructs.
The pattern coefficients showed that item 1 ("I am confident in my understanding of
Section 504 ofthe Rehabilitation Act of 1973") (.795), item 2 ("I am confident in my
understanding of the legal definition of disability") (.859), and item 3 ("I am confident in
my understanding of the Americans with Disabilities Act of 1990)" (.942) had large
coefficients on factor 1. After inspection of these items, I interpreted factor 1 to represent
the construct "Knowledge of Disability Law".
Item 22 ("Providing teaching accommodations to students with documented
disabilities is unfair to students without disabilities") (-.835), and item 24 ("I believe that
students with disabilities use the disability as an excuse when they are not doing well in
my class") (-.868) had large coefficients on factor 2. In addition, item 13 ("Providing
accommodations to students with disabilities is a way to ensure equal opportunity and
access to learning in higher education settings") (.425) and item 41 ("Making adequate
accommodations for students with documented disabilities in my courses is unrealistic
given time constraints and other job demands") (-.488) had moderate coefficients on
factor 2. After inspection of these items, I interpreted factor 2 to represent the construct
"Fairness".
Item 46 ("I include a statement in my syllabus inviting students with disabilities
to discuss their needs with me") (.726) had a large coefficient on factor 3, and item 47 ("I
make a statement in class inviting students with disabilities to discuss their needs with
49
me") (.584), item 51 ("I would like to modify curricular aspects of my course(s) in order
to create a more inclusive leaming environment for all students") (.565), item 52 ("I
prefer to use a variety of instructional formats in my class, including small group and
hands on activities") (.385), item 53 ("I believe it is my responsibility as an instructor to
provide consistent information and expectations on all assignments and exams") (.458),
and item 57 ("Students with disabilities should be able to perform just as well as students
without disabilities in my course[s]") (.416) all had moderate coefficients on factor 3.
Although initially six items loaded onto factor 3, four items were removed after applying
the decision rules. One item was removed because it cross loaded onto another factor,
and three others were not theoretically plausible with the other items. That is, when
grouped together, the items did not make logical sense as a measured construct.
Therefore, items 46 and 47 were retained for factor 3. After inspection of these items, I
interpreted factor 3 to represent the construct "Invitation to Disclose".
Items 14 and 27 ("I am willing to spend extra time (i.e., in addition to typical
office hours) meeting with students with documented disabilities to clarify and/or review
course related content or prepare for an upcoming exam" (.837) and "I am willing to
spend extra time (i.e., in addition to normal office hours) helping any student prepare for
an exam or review course material" (.867) respectively) had large coefficients on factor
4. After inspection of these items, I interpreted factor 4 to represent the construct
"Willingness to Invest Time".
Items 10 and 12 ("I am willing to provide copies of my lecture notes or outlines to
students with documented disabilities" (.790) and "I am willing to provide copies of my
50
overheads and/or PowerPoint presentations to students with documented disabilities"
(.790), respectively) had coefficients on factor 5. After inspection of these items, 1
interpreted factor 5 to represent the construct "Accessibility of Course Materials".
Item 15 ("1 am willing to allow a student with a documented disability to
complete extra credit assignments for academic success even when this option is not
stated on the course syllabus") (.925) had a large coefficient on factor 6, and item 16 ("1
am willing to allow any student to complete extra credit assignments in my course(s)")
(.690) had a coefficient on factor 6. After inspection of these items, 1 interpreted factor 6
to represent the construct "Willingness to Adjust Course Materials".
Items 38 and 39 ("Students with documented disabilities who request support
from Disability Services receive adequate services from that office" (.790) and "1 receive
adequate support from Disability Services to make appropriate accommodations for
students with documented disabilities" (.886), respectively) had large coefficients on
factor 7. After inspection of these items, 1 interpreted factor 7 to represent the construct
"Satisfaction with Campus Resources".
Finally, item 25 ("1 am willing to arrange extended time on exams for students
who have documented disabilities") (.941) had a strong coefficient on factor 8, and item
29 ("1 am willing to allow students with documented disabilities to take proctored exams
in a supervised location outside of the nOffi1al exam location") (.562) had a moderate
coefficient on factor 8. After inspection of these items, 1 interpreted factor 8 to represent
the construct "Willingness to Provide Testing Accommodations".
51
Although 27 total items were included in the EFA, eight of these items were
removed due to one or more of the stated decision rules. Therefore, the final model used
in the CFA included 19 items.
An oblique analysis was conducted and there was no decision made to rerun the
analysis with Varimax because there was a moderate amount of intercorrelation between
the factors. The highest intercorrelation was between factors 2 ("Fairness") and 8
("Willingness to Provide Testing Accommodations"), which was .491. The next highest
intercorrelation was between factors 2 ("Fairness") and 4 ("Willingness to Invest Time"),
which was .374. Finally, the weakest intercorrelation was between factors 2 ("Fairness")
and 6 ("Willingness to Adjust Course Assignments"), which was .004. Table 6 shows the
factor intercorrelations.
Descriptive statistics and factor reliability. Descriptive statistics and
Cronbach's a were examined for reliability, which addressed part of the first research
question. The response scale was a six-point level of agreement scale, ranging from 1
(strongly disagree) to 6 (strongly agree). Descriptive statistics for average scores are
reported for the entire survey and by subscale for ease ~f interpretation, as the eight
subscales were comprised of different numbers of items. The mean scores for the
subscales ranged from a high of 5.38 ("Willingness to Make Testing Accommodations")
to a low of2.99 ("Willingness to Adjust Course Assignments").
Table 6
Summary ofFactor Intercorrelations
Factor 1 2 3 4 5 6 7 8
1) Knowledge ofDisability Law 1.000
2) Fairness 0.243 1.000
3) Invitation to Disclose 0.335 0.185 1.000
4) Willingness to Invest Time 0.251 0.374 0.279 1.000
5) Accessibility of Course 0.158 0.205 0.291 0.293 1.000
Materials
6) Willingness to Adjust Course 0.063 0.004 0.109 0.143 0.156 1.000
Assignments
7) Satisfaction with Campus 0.118 0.233 0.173 0.149 0.067 -0.135 1.000
Resources
8) Willingness to Provide Testing 0.125 0.491 0.164 0.296 0.141 -0.155 0.159 1.000
Accommodations
VI
tv
53
A summary of reliability coefficients and descriptive statistics is shown in Table
7. Cronbach's a was examined for factor reliability. All factors had coefficient alphas of
.70 or higher, with the exception of factor 8, "Willingness to Provide Testing
Accommodations" (.64).
Table 7
Descriptive Statistics and Reliability Coefficients for Full ExCEL Survey and Subscales
Average Score
Measure Item n a Mean SD
ExCEL survey 19 .79 4.34 0.53
Knowledge of Disability Law 3 .86 3.01 1.31
Fairness 4 .77 5.02 0.76
Invitation to Disclose 2 .72 4.08 1.42
Willingness to Invest Time 2 .73 4.75 1.00
Accessibility of Course Materials 2 .74 5.09 1.00
Willingness to Adjust Course Assignments 2 .77 2.99 1.32
Satisfaction with Campus Resources 2 .80 4.36 0.85
Willingness to Make Testing Accommodations 2 .64 5.38 0.65
Confirmatory factor analysis. After conducting the EFA, the eight-factor structure was
tested with a confirnlatory model. As previously mentioned, eight items were removed
after the final EFA but prior to conducting the CFA due to cross loadings of .32 or
greater, failure to load onto any factor, or loading onto a low factor that was later
eliminated due to theoretical implausibility. The CFA model structure is shown in Figure
54
3. The results of the CFA yielded an inadmissible solution, which indicates that some
variance estimates are negative and suggests the model is inadequate or that the sample is
too small. More specifically, six ofthe eight factors in the model had only two items,
which typically indicates an underidentified model (Kline, 1998). There were two cases
of negative residual variance on two different factors: "Satisfaction with Campus
Resources" and "Willingness to Invest Time". Specifically, the items on these two
factors with the negative residual variance were item 14 ("1 am willing to spend extra
time (i.e., in addition to typical office hours) meeting with students with documented
disabilities to clarify and/or review course related content or prepare for an upcoming
exam") and item 39 ("1 receive adequate support from Disability Services to make
appropriate accommodations for students with documented disabilities"). The model was
tested again after eliminating these two factors, however, an inadmissible solution was
again the result. This time a negative residual variance occurred on another factor.
These findings show the factor structure ofthe instrument is not ready to be confirmed
yet. The results of the EFA are best utilized to inform decision making on item review
and development.
Figure 3. CFA model of ExCEL survey instrument
55
56
Group comparisons. To answer research questions two and three, I conducted
several Multivariate Analyses of Variances (MANOVAs) using the same sample from
the CFA model testing (n = 200). Specifically, research question two asked ifthere were
group differences according to gender, teaching status, and departmental affiliation. For
this analysis, gender and teaching status were specified as grouping variables and the
factor scores from the eight attitude and perception subscales obtained in the last EFA
were used as outcome variables. Due to insufficient sample size in some categories,
teaching status was converted from a variable with six categories into a dichotomous
variable. The variable was grouped dichotomously so that faculty were classified as (a)
tenure-line, or (b) nontenure, adjunct, visiting, or graduate assistant. I chose this
particular division because the original intent was to determine if there were differences
in attitudes and perceptions toward disability between faculty who were tenure-line and
those who were not.
Boxplots and histograms were examined for univariate and multivariate
nonnality, and no subscales had a significant deviation from normality. Because outliers
were present, the MANOVAs were conducted with and without the outliers. For the
gender and teaching status MANOVAs, there were no differences between the two
results, and therefore the outliers were not considered influential cases and left in the
analysis. For the interaction of gender by teaching status, there were differences between
results with and without outliers. Because of this difference, MANOVA results are
presented with the outliers excluded for the interaction of gender by teaching status.
Homogeneity of variance was tested using Box's Test of Equality of Covariance
Matrices. Examination of Box's M shows there was heterogeneity of variance, indicating
57
the observed covariance matrices of the dependent variables differ to a statistically
significant degree across groups, F(108, 38158) = 1.50,p = .001. Since heterogeneity of
variance was found, a more conservative a of .025 was used in the MANGVAs. An
estimate of effect size (TJp2) is also included in the presentation of results. Partial eta
squared (TJp2) is typically interpreted as the proportion of the total variance attributable to
a factor (Cohen, 1973; Levine & Hullett, 2002).
The results indicated a statistically significant difference in the multivariate
combination of the subscale scores based on teaching status, Wilks' A = .896, F(8, 189) =
2.74,p < .025, 'lp2 = .10. Results of the univariate tests indicated a statistically
significant difference in faculty attitudes and perceptions according to teaching status on
two ofthe eight subscales, specifically "Invitation to Disclose" (p = .002), and
"Willingness to Invest Time" (p = .011). However, there were no statistically significant
differences in the multivariate combination of subscale scores based on gender, Wilks' A
= .928, F(8, 189) = 1.82, p = .075, 'lp2 = .07, or the interaction of gender by teaching
status, Wilks' A = .923, F(8, 189) = 1.92, p = .052, 'li = .07 (with outliers included in
the analysis, results to follow for analysis excluding outliers).
In conjunction with conducting the MANGVAs, I examined the discriminant
function used to maximally differentiate between the groups for each subscale by
teaching status. The standardized discriminant function coefficients (SDFCs) and
structure coefficients (SCs) are reported in Table 8. There were only two levels of gender
and teaching status, and therefore only one discriminant function computed. The main
effect of teaching status on the eight subscales was significant and the canonical
58
correlation was .32, indicating that approximately 10% the variance was explained by the
discriminant function.
Table 8
Standardized Discriminant Function Coefficients (SDFCs) and Structure
Coefficients (SCs) for Each Subscale by Teaching Status
Subscale
Knowledge of Disability Law
Fairness
Invitation to Disclose
Willingness to Invest Time
Accessibility of Course Materials
Willingness to Adjust Course Assignments
Satisfaction with Campus Resources
Willingness to Provide Testing Accommodations
Teaching Status
SDFC SC
.335 .166
-.158 .445
-.468 -.172
.073 .138
.642 .512
-.606 -.439
.170 .324
.487 .564
The SDFCs indicated the "Accessibility of Course Materials" and "Willingness to
Adjust Course Assignments" subscales had the largest contribution to the formation of
the discriminant function, whereas the "Willingness to Invest Time", "Fairness", and
"Satisfaction with Campus Resources" subscales had the smallest contributions to the
formation of the discriminant function. The SCs indicated that "Willingness to Provide
Testing Accommodations", "Accessibility of Course Materials", and "Fairness" subscales
had the strongest correlations and were moderately related to the discriminant function,
59
demonstrated by correlations of.564 and .512. The subscales of "Willingness to Invest
Time", "Knowledge ofDisability Law", and "Invitation to Disclose", however, were
weakly related to the discriminant function, with correlations of .138, .166, and -.172
respectively. These results indicate tenure-line and nontenure faculty vary the most in
providing course materials in a variety of formats and allowing adjustments to course
assigmnents. As verified by the means, nontenure faculty are more likely to be flexible
in providing course materials in a variety of formats and allowing adjustments to course
assignments than tenure-line faculty. However, there is little variation between tenure-
line and nontenure faculty in attitudes regarding investing time toward helping students
outside of class time or posted office hours, ensuring equal treatment of students with and
without disabilities, and satisfaction with campus resources, such as the Disability
Services Office.
As mentioned, MANOVAs were conducted with and without the outliers in order
to determine if outliers should be considered influential cases. There were no differences
in results except for the interaction of gender by teaching status. With the outliers
excluded, results were statistically significant, Wilks' A = .919, F(8, 181) = 2.00,p
.048, 11p2 = .08. Results of the univariate tests indicated a statistically significant
difference according to the interaction of gender by teaching status on one subscale,
"Accessibility of Course Materials" (p = .028). These results should be interpreted with
caution, as they depend on the presence or absence of outliers.
In conjunction with conducting the MANOVA, I examined the discriminant
function used to maximally differentiate between the groups. The standardized
60
discriminant function coefficients (SDFCs) and structure coefficients (SCs) are reported
in Table 9.
Table 9
Standardized Discriminant Function Coefficients (SDFCs) and Structure
Coefficients (SCs) for Each Subscale by the Interaction ofGender by Teaching Status
Subscale
Knowledge of Disability Law
Fairness
Invitation to Disclose
Willingness to Invest Time
Accessibility of Course Materials
Willingness to Adjust Course Assignments
Satisfaction with Campus Resources
Willingness to Provide Testing Accommodations
Gender*Teaching Status
SDFC SC
-.335 -.315
-.526 -.062
.412 .401
.120 .089
.776 .544
-.237 -.127
-.485 -.435
.148 .156
There were only two levels of gender and teaching status, and therefore only one
discriminant function computed. The main effect of gender by teaching status on the
eight subscales was significant and the canonical correlation was .28, indicating that
approximately 8% the variance was explained by the discriminant function. The SDFCs
indicated the "Accessibility of Course Materials" subscale had the largest contribution to
the formation of the discriminant function at .776, with sizable contributions from the
"Fairness", "Satisfaction with Campus resources" and "Invitation to Disclose" subscales
61
as well. The "Willingness to Invest Time" and "Willingness to Provide Testing
Accommodations" subscales had the smallest contributions to the formation of the
discriminant function. The SCs indicated the "Accessibility of Course Materials", and
"Satisfaction with Campus Resources" subscales were moderately related to the
discriminant function, demonstrated by correlations of .544 and -.435. The subscales of
"Willingness to Invest Time", "Fairness", "Willingness to Adjust Course Assignments"
and "Willingness to Provide Testing Accommodations", however, were weakly related
to the discriminant function, with correlations of .089, -.062, -.127, and -.127
respectively. These results indicate the most variation occurred between faculty
according to gender and teaching status on the subscale "Accessibility of Course
Materials".
In order to completely address research question two, a second MANOVA was
conducted using college membership as the grouping variable. There were eight colleges
at the university, however two of the eight colleges had insufficient data, as MANOVAs
require at least as many participants per cell as there are dependent variables (Tabachnick
& Fidell, 2001). These two colleges, which were Honors College (n = 3) and Journalism
(n = 8), were eliminated from the MANOVA analysis. Additionally, although the Law
School was considered one of the eight colleges at the university, law school faculty were
not sent the survey because of procedural differences when handling disability
accommodation requests (see Chapter 2). Because of this, five separate colleges were
used for the college membership grouping variable, which included Arts and Sciences (n
= 90), Business (n = 22), Education (n = 35), Architecture and Allied Arts (n = 20),
62
Music and Dance (n = 21), and the eight attitude and perception subscales served as
outcome variables.
Boxplots and histograms were examined for univariate and multivariate
normality. No subscales had a significant deviation from normality. Outliers were
present, and MANOVAs were conducted with and without the outliers. There were no
significant differences between the results, and therefore the outliers were left in the
analysis and were not considered influential cases. Homogeneity of variance was tested
using Levene's Test of Equality of Error Variances and Box's Test of Equality of
Covariance Matrices. Examination ofBox's M shows there was homogeneity of
variance, indicating the observed covariance matrices of the dependent variables do not
differ to a statistically significant degree across groups, F(144, 18285) = l.34,p > .OOL
Levene's Test of Equality of Error Variances yielded non-significant results for all
subscales, indicating error variance of the dependent variables (subscales) does not differ
to a significant degree across groups.
The results indicated statistically significant differences in the multivariate
combination of the subscale scores based on college membership, Wilks' A = .528, F(8,
176) = 3.84,p < .001, 1lp2 = .14. Results of the univariate tests indicated a statistically
significant difference in faculty attitudes and perceptions according to college
membership on six of the eight subscales, specifically "Knowledge of Disability Law" (p
< .001), "Fairness" (p = .023), "Invitation to Disclose" (p <.001), "Accessibility of
Course Materials" (p = .002), "Willingness to Adjust Course Assignments" (p = .037),
and "Willingness to Make Testing Accommodations" (p = .023).
63
In conjunction with conducting the MANGVA, I examined the discriminant
function used to maximally differentiate between the groups. There were four
mathematically possible discriminant function roots to differentiate between the five
groups, but only two ofthe four roots were statistically significant, which means there
were two possible weighted linear combinations that describe the differences among
faculty according to college membership. The canonical correlation for the first root was
.59, indicating that approximately 35% ofthe variance in college membership was
explained by the discriminant function, and the canonical correlation for the second root
was .38, indicating approximately 14% of the variance in college membership was
explained by the second root. The standardized discriminant function coefficients
(SDFCs) and structure coefficients (SCs) are reported in Table 10. The SDFCs indicated
the "Knowledge ofDisability Law" and "Invitation to Disclose" subscales had the largest
contribution to the formation of the first root, whereas the "Willingness to Invest Time",
"Fairness", and "Willingness to Adjust Course Assignments" subscales had the smallest
contributions to the formation of the first root. The SCs indicated that the "Knowledge of
Disability Law" and "Invitation to Disclose" subscales were strongly related to the first
root demonstrated by correlations of .742 and .633. The subscales of "Willingness to
Provide Testing Accommodations" and "Satisfaction of Campus Resources", however,
were weakly related to the first root, with correlations of .041 and -.106, respectively.
This finding indicates the most variation in faculty attitudes across colleges occurred in
their ways of inviting students to disclose disabilities, as well as understanding and
knowledge oflegal mandates related to disability.
64
Table 10
Standardized Discriminant Function Coefficients (SDFCs) and Structure
Coefficients (SCs) for Each Subscale by College Membership
Root 1 Root 2
Subscale SDFC SC SDFC SC
Knowledge of Disability Law .716 .742 .335 .166
Fairness .117 .221 -.158 .445
Invitation to Disclose .547 .633 -.468 -.172
Willingness to Invest Time -.072 .255 .073 .138
Accessibility of Course Materials .272 .278 .642 .512
Willingness to Adjust Course Assignments .125 .192 -.606 -.439
Satisfaction with Campus Resources -.205 -.106 .170 .324
Willingness to Provide Testing Accommodations -.151 .041 .487 .564
For the second root, the SDFCs indicated that the "Accessibility of Course
Materials" and "Willingness to Adjust Course Assignments" subscales had the largest
contribution, whereas the "Willingness to Invest Time", "Fairness" and "Satisfaction of
Campus Resources" subscales had the smallest contributions to the formation of the
second root. The SCs indicated that the "Willingness to Provide Testing
Accommodations" and "Accessibility of Course Materials" subscales were strongly
related to the second root with correlations of .564 and .512, respectively. The subscales
of "Willingness to Invest Time" and "Knowledge ofDisability Law", however, were
weakly related to the second root, with correlations of .138 and .166, respectively. This
finding indicates the most variation in faculty attitudes across colleges occurred in their
65
willingness to use inclusive instructional practices, which entail flexibility in required
course assignments and the use of accessible course materials.
The third research question asked whether faculty members with prior experience
in disability awareness training have more positive attitudes and perceptions than those
who have not received training. To answer this question, another MANOVA was
conducted where I grouped faculty dichotomously according to whether or not they had
participated in prior disability-focused training experiences. "Training" was determined
based on respondents' answers to a series of questions in the beginning of the survey.
First, respondents were asked if they had received any prior disability-focused training, to
which they answered Yes or No. A follow-up question asked what type of training, and
the options included (a) attended a workshop (b) took one or more courses, (c) read books
or articles, (d) visited website(s), and (e) other. Overall, 29% of faculty had participated
in prior training (n = 57) and 71 % had not participated in any form of prior training (n =
143). The eight attitude and perception subscales were again used as dependent variables
for this analysis.
Boxplots and histograms were examined for univariate and multivariate
normality. No subscales had a significant deviation from normality. Outliers were
present, and MANOVAs were conducted with and without the outliers. There were no
significant differences between the results, and therefore the outliers were not considered
influential cases and left in the analysis. Homogeneity of variance was tested using Box's
Test of Equality of Covariance Matrices. Examination of Box's M shows there was
heterogeneity of variance, indicating the observed covariance matrices ofthe dependent
variables differ to a statistically significant degree across groups, F(36, 41276) = 1.99,p
66
< .001. Since heterogeneity of variance was found, a more conservative a of .025 was
used in the MANGVA.
The between subjects multivariate results indicated a statistically significant
difference in the multivariate combination of the subsca1e scores based on prior training
experiences, Wilks' A = .800, F(8, 191) = 5.96,p < .001, 11p2 = .20. Results of the
univariate tests indicated that faculty who had experienced prior training had higher
scores on the subsca1es measuring Knowledge of Disability Law (p < .001) and
providing students with Invitation to Disclose (p < .001) than did faculty who had not
previously attended disability-focused training.
In conjunction with conducting the MANGVA, I examined the discriminant
function used to maximally differentiate between the groups. The SDFCs and SCs for
each subsca1e are included in Table 11. There were only two levels of prior training
experience and therefore only one discriminant function. The canonical correlation was
.44, indicating that approximately 19% of the variance was explained by the discriminant
function. The SDFCs indicated that the "Knowledge of Disability Law" (-.837) had the
largest contribution to the formation of the discriminant function, and "Invitation to
Disclose" also had a moderately large contribution (-.438), However, the "Accessibility
of Course Materials", "Willingness to Adjust Course Assignments", "Willingness to
Provide Testing Accommodations" and "Satisfaction with Campus Resources" subscales
had the smallest contributions to the formation of the discriminant function. The SCs
indicated that the "Knowledge of Disability Law" and "Invitation to Disclose" subsca1es
were strongly related to the discriminant function demonstrated by correlations of -.877
and -.559. The subsca1es "Accessibility of Course Materials"," Willingness to Adjust
67
Course Assignments"," Willingness to Provide Testing Accommodations" and
"Satisfaction with Campus Resources", however, were weakly related to the discriminant
function, with correlations of -.075, .069, -.036, and -.100, respectively. This finding
indicates the most variation occurs between faculty with and without prior training in
their current practices inviting students to disclose disabilities and their knowledge of
legal mandates related to disability.
Table 11
Standardized Discriminant Function Coefficients (SDFCs) and Structure
Coefficients (SCs) for Each Subscale by Prior Training
Subscale
Knowledge of Disability Law
Fairness
Invitation to Disclose
Willingness to Invest Time
Accessibility of Course Materials
Willingness to Adjust Course Assignments
Satisfaction with Campus Resources
Willingness to Provide Testing Accommodations
Training
SDFC SC
-.837 -.877
-.121 -.193
-.438 -.559
.181 -.121
-.006 -.075
.156 .069
-.107 -.100
.057 -.036
68
CHAPTER IV
DISCUSSION
In this study, I sought to examine the psychometric properties of the ExCEL
Survey, an instrument that purports to measure the attitudes and perceptions of university
facuIty toward students with all disability types. These attitudes and perceptions were
centered on the willingness to accommodate students with disabilities and to adjust
instruction and course assignments to incorporate the tenets of Universal Design. In this
chapter, I will summarize and interpret the findings of this study that pertain to each
research question. As well, I will discuss limitations, address implications of the findings
as they relate to applied settings, and discuss pertinent directions of future research.
Factor Structure
Prior to conducting the study, items were written intending to capture eight
hypothesized factors. The final subscales that emerged from the EFA differed somewhat
from the hypothesized factors and included: "Knowledge of Disability Law", "Faimess",
"Invitation to Disclose", "Willingness to Invest Time", "Accessibility of Course
Materials", "Willingness to Adjust Course Assignments", "Satisfaction with Campus
Resources", and "Willingness to Make Testing Accommodations". This factor structure
explained 67% of the total variance of the original items. Item pattem coefficients on
each subscale ranged from .94 to .38. The mean scores for the subscales ranged from a
high of 5.38 ("Willingness to Make Testing Accommodations") to a low of 2.99
69
("Willingness to Adjust Course Assignments"), indicating that participants were most
likely to agree to meet requests from students regarding testing accommodations, but
least likely to adjust course assignments at the request of students.
After three EFAs were conducted, 30 items were removed. The decision to
remove items from each EFA was based on the decision rules outlined in Chapter 3. The
most common reason for removing items was cross loading onto two or more factors. In
fact, of the 30 removed items, 23 (77%) were removed because they cross loaded onto
two or more factors. This finding suggests the hypothesized constructs overlapped too
much; in other words, items that were written with the intent to capture one construct
overlapped with another construct, so that the items were not worded specifically enough
for each construct. These results suggest the factor structure of the instrument is not
ready to be confirmed, and the focus should be on further item development. Through
the results of this study, it is now clear which items load onto multiple factors and
therefore did not function as intended. These items need to be re-written so they are
more focused and specific to the intended construct and less general to other constructs,
and tested again so cross loadings will be avoided in future analyses.
The factor structure of the instrument was not confirmed, as verified by the
inadmissible solution resulting from the CFA. Although it is adapted from a previous
instrument (Murray et aI., 2008), it is important to consider the preliminary stages of
development for this instrument. In this stage of development, some items were
removed, rewritten, and one construct was added (including the addition of9 new items).
These are major changes to the instrument and more than one field-test will be necessary
to confirm the factor structure. Further item development is needed prior to conducting
70
another CFA. Additionally, there is no evidence a CFA was conducted on the previous
instrument, so the inadmissible solution was not necessarily due to the new items.
There were some differences between the hypothesized and actual factors that
warrant further discussion. Specifically, these differences involve three of the eight
hypothesized factors, measuring knowledge of disabilities (both visible and non-visible),
willingness to provide accommodations, and willingness to create more inclusive
learning environments (essentially to adapt instruction based on the tenets of Universal
Design). Of the 30 removed items, 14 (47%) were items intended to capture these three
hypothesized constructs. Further, these are three crucial facets in measuring faculty
attitudes and perceptions toward disability. Because these findings indicate these
constructs were not measured in the ExCEL survey, it is particularly important to further
examine the removed items and determine how they can be revised in future attempts to
measure these constructs.
Willingness to accommodate. The "Willingness to Accommodate" construct is
a cornerstone of the instrument, as much of postsecondary students' experiences involve
their interaction with faculty members when requesting accommodations. For most
students with disabilities, accommodations area crucial component of success, and being
denied accommodations could be a major factor in student outcomes. In previous studies
where similar instruments were examined, this construct was present (Murray et aI.,
2008; Vogel et aI., 2008). Murray et ai. (2008) found their instrument had three different
factors directly related to accommodations: "Willingness to Provide Major
Accommodations", "Willingness to Provide Teaching Accommodations", and
71
"Willingness to Provide Testing Accommodations". Vogel et al. (2008) found one
general construct that related to knowledge of accommodations.
Similar to Murray et al. (2008), the results of this study indicated the presence of
one factor specifically related to providing accommodations, which was "Willingness to
Provide Testing Accommodations". There were two other factors related to
accommodating students, "Willingness to Adjust Course Assignments" and
"Accessibility of Course Materials", both of which included items related to providing
accommodations. For example, one of the items of "Accessibility of Course Materials"
is "I am willing to provide copies of my lecture notes to students with documented
disabilities", a typical accommodation request from students with learning disabilities,
Attention Deficit/Hyperactivity Disorder, or students with visual impairments.
Therefore, three factors emerged representing the "Willingness to Accommodate"
construct, which specifically addressed faculty attitudes and perceptions of (a) providing
testing accommodations to students, such as extended time on exams, (b) changing
course assignments, such as adding an extra credit assignment, or (c) allowing access to
course materials that are not typically shared, such as lecture notes. All of these actions,
if taken by a faculty member, would be seen as a type of accommodation for students
with disabilities, and it is therefore important to emphasize that three of the eight factors
on the instrument measure "Willingness to Accommodate".
In my hypothesized factors, I intended to capture attitudes and perceptions toward
students with different disability types, specifically visible and non-visible disabilities.
Particularly, there were two items written to measure this distinction, which were (a) "I
have provided recommended accommodations (e.g. extended time, private testing room,
72
etc.) for students with leaming disabilities in my course(s)" and (b) "I have provided
recommended accommodations (e.g. large-print materials, captioned videos, etc.) for
students with disabilities in my course(s)". The first item lists examples of typical
accommodation requests from students with leaming disabilities, which are considered
non-visible. The second item lists examples of typical accommodation requests from
students with more obvious, visible disabilities. Both of these items were deleted from
the instrument after conducting the first EFA because they cross-loaded onto two factors
with coefficients of .32 or greater. This finding suggests the items may not have been
written specifically enough to capture the different attitudes and perceptions faculty
members have toward students with visible and non-visible disabilities. However, there
are several problems with these items, which may have contributed to their removal from
the instrument. First, both items began with the stem "I have provided accommodations",
which measures whether faculty members have done this in the past, not necessarily
whether they would provide the accommodation if a student requested it. In hindsight,
the stem should have been written as "I am willing to provide" which captures all faculty
members' willingness, essentially the attitude I intended to measure. Second, because the
items loaded onto two factors, the hypothesized constructs were not captured specifically
enough so that they represented different, although relative, facets of faculty attitudes and
perceptions toward disability. This finding may suggest that since disability types are so
dynamic and accommodation requests typically differ to a great degree, one construct to
encompass all is not theoretically plausible.
73
Universal design. Little to no research exists on quantitative attempts to measure
faculty attitudes and perceptions toward adopting the tenets of Universal Design into their
own instructional practices. While faculty attitudes and perceptions toward disability
have been previously measured, none of those instruments included a component to
address tenets of Universal Design. This particular portion of the instrument tested
uncharted territory. The tenets of Universal Design were mentioned in Chapter 1 (see
Table 1, page 15).
Nine items were written to measure the construct "Willingness to Create Inclusive
Learning Environments", a construct intended to capture the instructional aspects of
Universal Design. Seven ofthe nine items were removed after the first or second EFA
because they either (a) cross loaded on two or more factors with pattern coefficients of
.32 or greater, (b) did not load onto any factors, or (c) did not load along with other items
to create a theoretically plausible construct. Like the "Willingness to Accommodate"
construct, these results indicate the items were not written specifically enough to
uniquely capture inclusive instruction and therefore Universal Design. These results are
not unexpected for such a newly developed instrument, especially given the little
empirical evidence behind the theoretical framework of Universal Design. Two items, on
the other hand, were retained after the final EFA was conducted, but these items did not
load onto one factor together that represented "Willingness to Create Inclusive
Instructional Environments". Instead, the two items loaded onto different factors, which
were "Accessibility of Course Materials" and "Willingness to Adjust Course
Assignments". Recall these same two factors mentioned in the discussion on the
"Willingness to Accommodate" construct. These findings show a large amount of
74
overlap between the intended constructs of "Willingness to Accommodate" and
"Willingness to Create Inclusive Learning Environments", a finding that illustrates the
inherent overlap between accommodating students and adopting the principles of
Universal Design. Ketterlin-Geller and Johnstone (2007) surmised that faculty members
are less likely to make individual accommodations for students should they choose to
embed LID principles into their course framework. These findings support this theory
because so many items cross loaded, showing the degree to which both constructs
overlap. More research must be conducted to determine greater distinctions between a
faculty member's willingness to accommodate and willingness to modify instruction. It
seems there would be different attitudes toward them, as willingness to modify
instruction could be more time-consuming for faculty members. However, further
research must be done confirm and disconfirm these theories.
Overall, these findings show measuring faculty attitudes and perceptions of
disability accommodations and Universal Design is important. The accommodations-
related constructs overlapped with the Universal Design constructs, and some items
intended for one construct resulted on another construct. This finding suggests there is a
good amount of overlap between attitudes and perceptions toward accommodation
requests and Universal Design principles. Further research of this nature is needed where
items are more fully developed that represent each construct more holistically and
discriminantly. In particular, future studies should measure both constructs and compare
faculty attitudes toward making accommodation requests and modifying instruction
according to the tenets of Universal Design.
75
Reliability
Reliability of the instrument was assessed using Cronbach's a, which was
examined within factors and the entire instrument. The overall reliability was strong, a =
.79, approaching the preferable .80 criteria and exceeding the acceptable .70 criteria
(Nunnally, 1975). Reliability within subscales ranged from highest "Knowledge of
Disability Law" (a = .86) to lowest "Willingness to Make Testing Accommodations" (a
= .64). "Willingness to Make Testing Accommodations" was the only subscale that fell
below the acceptable.70 criteria. However, it is important to note the only measure of
reliability was internal consistency, and no alternate forms or test-retest design was
utilized simply due to the sampling plan. Measures of internal consistency (such as
Cronbach's a) provide the degree to which the items on the instrument combine to
measure a single trait; however, Cronbach's a is not the most robust form ofreliability
(Henson, 2001).
Validity
The recommended steps to establishing construct validity were followed in this
study, including the identification of a theoretical model, development of an initial item
pool, and item selection and psychometric evaluation (AERA, APA, & NCME, 2009;
Clark & Watson, 1995). Partial evidence of construct validity was obtained through the
findings in this study, particularly evidence of content, convergent, and discriminant
validity.
Content validity was established in several ways: (a) many of the items were
taken from a pre-existing survey instrument that showed good promise of reliability and
validity (Murray et aI., 2008), and (b) of the items that were newly added, all content was
76
based on an extensive literature review of Universal Design, specifically the tenets
related to instruction. In addition, the items were reviewed by the Director of Disability
Services to ensure consistency with their understanding of Universal Design. Content
validity of the instrument will be strengthened by using the outcomes of the item analysis
to drive future instrument development. As previously mentioned, many items were
omitted after three rounds ofEFA. A logical next step is to carefully examine those
deleted items and detennine if the problem lies in the wording of the item, or if the
content is not necessary for the intended constructs.
The eight factor structure of the ExCEL survey shows evidence of convergent and
discriminant validity. More specifically, items grouped together represent a set of
indicators with moderate to strong intercorrelations within each construct, which is
evidence of convergent validity. Because the factor correlations were weak to moderate,
there is evidence of discriminant validity (Kline, 1998). As previously stated, because so
many items cross loaded on one or more factor, future item development must focus on
the distinction between the constructs. For example, in drafting new items, researchers
should ask: what distinguishes a faculty member's willingness to provide testing
accommodations from willingness to adjust course assignments?
The results of the CFA provide evidence against construct validity, as the factor
structure could not be confirmed. Additionally, the presence of two-item subscales
suggests the model may be incorrect or flawed, and more items are needed to adequately
measure these constructs. Therefore, the results of this study show partial evidence for
and against the validity of the instrument. As mentioned, the instrument has substantial
reliability evidence. A single administration of a version of the instrument cmmot
77
provide overall definitive evidence the ExCEL survey measures precisely what it
purports to measure (Messick, 1995). Therefore, it is especially important to focus on
item development as the most immediate next step in future research. These findings
show the instrument has great promise, but must undergo multiple iterations in order to
show strong evidence of construct validity.
In evaluating evidence for external validity, it's important to consider several
limitations to the study. One aspect of external validity is the extent to which results can
be generalized from sample to population, and therefore the nature of the sample is
important to consider. The response rate was lower than anticipated (27% overall)
although adequately representative of the faculty population at this particular university
setting. Findings should be interpreted with caution due to the low response rate. Also,
the survey was voluntary. Only faculty who desired to volunteer their time responded,
potentially leading to a sampling bias.
Group Differences
Research questions two and three specifically addressed group differences on the
measure. To examine whether there were group differences on the measure, based on
gender, teaching status, and college membership, two separate MANOVAs were
conducted using the eight subscales as dependent variables. To examine whether there
were group differences in the measure based on prior disability-focused training
experiences, I conducted another MANOVA using the eight subscales as dependent
variables. Please note, two separate MANOVAs were conducted in order to answer
research question two; the first MANOVA examined group differences based on gender
and teaching status, and the second MANOVA was based on college membership. I
78
chose this design because the sample size of each group in college membership was not
adequate enough to include three predictors in the same MANOVA.
The results of these analyses indicated there were no statistically significant
differences in subscale score based on gender, which contradicts earlier findings that
female faculty members are more likely to meet accommodation requests from students
with disabilities (Leyser et aI., 1998; Murray et aI., 2008; Skinner, 2007). On the other
hand, there were statistically significant differences in subscale score based on teaching
status, where tenure-line faculty members scored significantly lower overall than
nontenure faculty members. These findings suggests that nontenure faculty members are
more likely to meet accommodation requests and adopt inclusive instructional practices
than tenure-line faculty members, which is inconsistent with previous findings, where
faculty rank did not have a significant effect on attitudes and perceptions (Rao, 2002).
There is no specific explanation why these results differed from previous studies, except
to consider the different university contexts as a major factor in influencing faculty
attitudes and perceptions toward disability. University context may be more influential
than gender or teaching status. Univariate tests in teaching status revealed statistically
significant differences in tenure-line faculty and nontenure faculty on two of the eight
subscales, which were "Invitation to Disclose" and "Willingness to Invest Time". These
findings suggest that tenure-line faculty are less likely to make an announcement in class
or include a statement of their syllabus that invites students with disabilities to disclose to
them at the beginning of the academic term. Also, they are less likely to spend extra
time, outside of typical office hours, helping students with disabilities with class
assignments or exam preparation.
79
These results were paralleled by the examination of which factors were correlated
with the discriminant function. These findings suggest the greatest amount of variation in
attitudes and perceptions between tenure-line and nontenure faculty members is on the
"Accessibility of Course Materials" and the "Willingness to Adjust Course Assignments"
subscales, which is interesting given these subscales consist of items that measure
flexibility and accessibility, more of the inclusive instructional practices and less of the
legal mandates associated with accommodations. These subscales show the most
amount of variation, indicating nontenure faculty are more likely to be flexible and adopt
inclusive instructional practices than tenured/tenure track faculty.
In order to test whether the outliers should be considered influential cases, the
MANGVA was conducted again with the outliers excluded. Results showed statistically
significant subscale scores based on the interaction of gender by teaching status. After
examination of the univariate tests and paralleled by the results of the discriminant
function, the most amount of variation between groups occurred on the "Accessibility of
Course Materials" subscale, where nontenure male faculty were most likely to provide
accessible course materials than nontenure female, tenure-line male, and tenure-line
female faculty. However, these results should be interpreted with caution because they
depend on the presence or absence of outliers, and were not significant when outliers
were included in the analysis.
The second MANGVA examined group differences according to college
membership, and answered the second part of research question two, specifically
addressing group differences in subscale scores based on membership in the various
colleges within the university. As previously mentioned, this analysis was limited to five
80
of the eight colleges (two colleges had inadequate sample sizes and one college was not
included in the study). The results of this analysis showed statistically significant
differences in subscale score based on college membership. This finding suggests faculty
attitudes and perceptions toward disability varied between the colleges on campus, and
attitudes and perceptions may be influenced depending on which college a faculty
member belongs to, a finding consistent with previous studies (Leyser et aI., 1998;
Murray et aI., 2008; Rao, 2002; Skinner, 2007).
More specifically, on the "Knowledge of Disability Law" subscale, faculty
members in Education scored significantly higher than faculty from the other four
colleges, indicating a greater degree of knowledge oflegal mandates around disability.
Education faculty scored the highest on the "Fairness", "Invitation to Disclose",
"Accessibility of Course Materials" subscales as well. However, faculty from the
College of Business scored the highest on the "Willingness to Make Testing
Accommodations" subscale. These findings are generally consistent with previous
findings that faculty attitudes and perceptions differ according to college or departmental
affiliation, and that Education faculty tend to be the most accommodating and exhibit the
most amount of fairness toward students with disabilities (Leyser et aI., 1998; Murray et
aI., 2008; Skinner, 2007). However, there are no previous findings that Business faculty
members are particularly accommodating, and this finding is therefore inconsistent with
the literature.
More investigation is warranted to further understand why Business faculty
members scored particularly high in their attitudes and perceptions toward providing
exam accommodations to students with disabilities, and why Education faculty members
81
have a greater knowledge of disability legal mandates, have accessible course materials,
exhibit fairness toward students with disabilities, and make an outward invitation to
students to disclose their disability. Potentially, this finding is a result ofpractices
specific to the Colleges ofBusiness and Education, where perhaps College-wide policies
and procedures about disability contributed to the differences, but more investigation is
needed in order to clarify.
The second root, or weighted linear combination, showed "Accessibility of
Course Materials", "Willingness to Adjust Course Assignments", and "Willingness to
Provide Testing Accommodations" had the largest contribution to the formation of the
discriminant function, indicating the most amount ofvariation in attitudes and
perceptions between faculty from the five colleges occurred in these subscales. Again, as
in the weighted linear combination of the tenure-line versus nontenure faculty
MANOVA, the most amount of variation occurred in subscales measuring flexibility in
instructional methods and course assignments. This finding shows that faculty members
in the five colleges differ in their attitudes and perceptions to a significant degree when it
comes to flexibility and adopting inclusive instructional practices, which is not
particularly surprising considering there are no policies in place where faculty are
required to adopt such practices.
Finally, a third MANOVA was conducted in order to address research question
three, specifically regarding differences between faculty who have and have not received
prior disability-focused training. The results of this analysis showed statistically
significant differences in subscale score based on prior disability-focused training or
experiences. Essentially, faculty who receive training opportunities are more likely to
82
understand and carry out legal responsibilities affiliated with disability and are more
likely to make a statement in class or on their syllabus that encourages students to
disclose their disability. These findings suggest that university faculty training could
playa crucial role in ensuring students with disabilities receive a quality higher education
experience. Further, these findings suggest that as long as faculty are made aware of the
laws and strategies to accommodate students with disabilities, they are likely to comply
and willing to help. Faculty members who received prior disability-focused training,
however, were not necessarily likely to adopt inclusive instructional practices or provide
accessible course materials. A potential next step is to consider the content or structure
of training opportunities to determine if they centered on laws and accommodations. If
this is the case, future research studies should examine the effect of training opportunities
based on the tenets of Universal Design in order to determine if faculty members tend to
modify their instruction after learning how to create accessible and inclusive learning
environments.
Limitations
There are several limitations to this study that should be considered in the
interpretation of these findings. First, there is potential sampling bias. The factor
structure was examined using a sample of faculty from a single university in the Pacific
Northwest. Further, the survey was voluntary so the sample included only those faculty
members willing to invest their time. All faculty members from the initial sample (N =
1084) were given an incentive, a $3 coffee coupon, but this may not have been all that
influential. The response rate was far from ideal, at 27%, although it is consistent with
other studies involving university faculty and disability awareness (Bourke et al., 2000;
83
Murray et aI., 2008; Murray et aI., 2009; Skinner, 2007; Vogel et aI., 1999, Vogel el aI.,
2008), and suggests that, perhaps, there are challenges to sampling university faculty.
Regardless, factor analysis is sample dependent, and it is possible a different factor
structure would have resulted had the response rate been higher. Also, it's important to
note that this particular university has a rather high percentage of white faculty (82%)
compared to all ethnic minority faculty (13%) with 5% of the faculty declining to report.
A second limitation is potential respondent bias. The instrument was a self-report
survey, which allows the potential for dishonest responses. Due to the sensitive nature
of the item content, it is likely participants desired to give the most politically correct
response, even if it was not the most honest response. This bias was avoided by assuring
confidentiality to participants, however, the potential for respondent bias is important to
consider in interpreting study results.
A third limitation is the case-to-variable ratio. In EFA, the recommended case-to-
variable ratio is 5:1 (Costello & Osborne, 2005; Fabrigar et aI., 1999; Stevens, 2002), and
this study had a 2:1 ratio. In the effort to complete the cross-validation study with both
an EFA and CFA, the ratio was compromised so there would be a larger sample of 200
for the CFA. In the end, I determined further item development is needed prior to
conducting another CFA.
A fourth limitation, and essentially the reason why the CFA solution was
inadmissible, was the presence of two-item subscales. Six of the eight subscales included
only two items and while two items per factor might be acceptable, a minimum of three is
recommended, four is more desirable, and five or more strong items (loadings .50 or
higher) indicate a strong factor (Costello & Osborne, 2005; Kline, 1998). Oftentimes,
84
two-item subscales are unstable, and may lead to problems such as underidentification
and nonconvergence of the measurement model. Two-item subscales were not intended,
and all hypothesized factors had a minimum of 6 items prior to conducting analyses.
However, so many items were removed after the three rounds ofEFA, resulting in item
removal and subsequent two-item subscales. This limitation shows the importance of
item development as the crucial next step in further development of the ExCEL survey.
Careful time must be spent on item development prior to conducting another CFA.
Factors should include more than enough items so that items can be removed but will not
result in two-item subscales. More discrimination between factors is necessary to avoid
cross loadings.
Finally, the quantitative instrument poses some limitations. Although the results
provided an overall sense of faculty members' perceptions and attitudes toward disability,
much ofthe detail of these attitudes was not captured. It is important to further explore
dominant themes found in the findings through qualitative research methods. Focus
groups and individual interviews with faculty members, students with disabilities, and
Disability Services personnel on very targeted and focused topics may provide rich and
detailed findings that will not be captured with a quantitative instrument. This instrument
provides the specific areas on which to focus, but qualitative research is needed for
further, more in-depth exploration of these areas.
Implications for Practice
Despite these limitations, the ExCEL survey shows evidence of initial reliability
and validity and therefore can be used as a tool for assessing university faculty attitudes
and perceptions of disability. Specifically, university administrators may use subsca1e
85
scores to examine willingness to provide accommodations and to adjust instruction to
include more inclusive principles associated with Universal Design. As previously
mentioned, it is becoming more common for effective teaching to be an area of
evaluation included in faculty performance reviews. This shift may result in an increased
demand for postsecondary faculty members to participate in professional development,
particularly in the areas of disability awareness and inclusive instruction that will benefit
all students. Survey results can be used to plan for targeted faculty professional
development trainings at a college or departmental level. Mean scores on subscales will
inform university administrators of areas of weakness in particular departments, among
certain faculty ranks, or within the entire university. For example, the survey results
showed a particularly high level of awareness of "Knowledge of Disability Law" among
Education Faculty. In recruitment efforts, university administrators know that the
Education faculty need less training about legal mandates, and more training in inclusive
instructional principles. Trainings could be targeted by department, and Department
Heads and Deans may use this information to plan for the most effective training in areas
of weakness particular to their department or college.
Further, the ExCEL survey has the potential to be used by faculty as a self-
assessment of disability awareness. Faculty members could take the ExCEL survey
online and receive immediate feedback on results, highlighting strengths and weaknesses
and recommendations for improvement. These recommendations may include tips for
providing accommodations and incorporating more inclusive instructional principles into
course planning and delivery. Ideally, faculty should be able to use their results to
improve their instruction in order to reach a wider range of diverse learners.
86
It is most important to recall the larger picture: more students with disabilities
pursue postsecondary education now than ever before. Although few pursue education at
four-year universities, the number continues to increase. With this change, disability
awareness must be increased among university faculty, who tend to have the most
amount of direct contact with and influence on the overall postsecondary experiences of
students with disabilities. Further, following the tenets of Universal Design will promote
the use of inclusive instructional practices among all faculty. This change will benefit
not only students with disabilities, but all university students, especially given the
increase in diverse learners on college campuses nationwide. This study shows the
crucial first step to assessing any institutional climate is to reliably and validly measure
faculty attitudes and perceptions toward students with disabilities. Such assessment can
and will lead to targeted faculty interventions that will enhance the overall quality of
education received by all students, with and without disabilities.
APPENDIX A
EXCEL SURVEY
Project Excel: Faculty Survey
a. Gender
87
Female 0 Male 0
b. Teaching Status (check one)
Tenured 0
Tenure Track 0
Non-tenure Track 0
Adjunct 0
Visiting 0
Post-retire 0
Graduate assistant 0
c. Rank (check one)
Instructor 0
GTF 0
Assistant Professor 0
Associate Professor 0
Full Professor 0
d. University of Oregon College (check one)
Arts and Sciences 0 Business 0
Architecture and Allied Arts 0
Honors CollegeD
Education 0
Music and Dance 0
Law 0
Journalism 0
e. Arts and Sciences only (please check division):
Humanities 0 Social Sciences 0 Sciences 0
88
f. How many years have you been teaching at the postsecondary level? _
g. I teach primarily (check one)
o General education courses
o Elective courses
o Discussion/Lab section (subsection oflecture)
o Major-specific courses
o Graduate level courses
h. In the past five years, how many college students with disabilities have you
taught or worked with?
o None
o 1-5
o 6-10
o 10-20
o Over 20
o Don't know/not sure
i. What is your age? _
j. Have you had any personal experiences with disability? (please check all that apply)
D I have a disability
D Family member, friend or other personal contact has a disability
D I have worked with or taught students with disabilities
D No, I have not had any personal experiences with disability
k. If you were to attend a training session at University of Oregon, which topics
would you find most relevant and/or interesting? Rank order topics (1= most
important, 5=least important)
D Increasing my understanding of disability issues in the college setting
Dlncreasing my understanding of student experiences
DLeaming more about inclusive instructional approaches that would reduce the need for
student-specific accommodations
DBetter understanding of Disability Services and how they can support me and my
students
DIn-depth understanding of specific types of disabilities
Directions: The following questions pertain to training you may have received in
supporting students with disabilities.
la. Have you ever received training related to disability or working with
college students with disabilities?
Yes No
o D
If no, please skip to Survey Directions on the next page. If yes, please
answer the following:
lb. What type of training? (please check all that apply)
D Attended a workshop
D Took one or more courses
o Read books or articles
D Visited Website(s)
D Other (please describe or list below):
I c. How much training?
D Less than 1 hour
D between 1 and 3 hours
D between 4 and 6 hours
D between 7 and 9 hours
o 10 hours or more
Continue survey on next page ....
89
Survey Directions: Please rate the following 60 items from 1 (Strongly Disagree) to 6
(Strongly Agree)
Response Format
90
1 = Strongly
Disagree
2 = Disagree 3 = Somewhat 4 = Somewhat
Disagree Agree
5 = Agree 6 = Strongly
Agree
1. I am confident in my understanding of section 504 102030405060
ofthe Rehabilitation Act of 1973.
2. I am confident in my understanding of the legal 102030405060
definition of "disability"
3. I am confident in my understanding of the 102030405060
Americans with Disabilities Act (1990).
4. I believe that students with learning disabilities can 102030405060
be successful at the university level.
5. Students with non-visible disabilities (e.g. learning 102030405060
disability, psychological condition, etc.) are often
reluctant to disclose their disability.
6. I would like more information about the needs of 102030405060
students with disabilities at the University of
Oregon.
7. I am sensitive to the needs of students with 102030405060
disabilities at the University of Oregon.
8. Students with disabilities are able to compete 102030405060
academically at the university level.
9. Students with disabilities attend postsecondary 102030405060
schools at rates proportionate to students without
disabilities.
91
10. I am willing to provide copies of my lecture notes 102030405060
or outlines to students with documented disabilities.
11. I am willing to provide additional time to complete 10 20 30 40 50 60
assignments in my course(s) to any student who
expresses a need.
12. I am willing to provide copies of my overheads 102030405060
and/or PowerPoint presentations to students with
documented disabilities.
13. Providing accommodations to students with 102030405060
disabilities is a way to ensure equal opportunity and
access to learning in higher education settings.
14. I am willing to spend extra time (i.e., in addition to 102030405060
typical office hours) meeting with students with
documented disabilities to clarify and/or review
course related content or prepare for an upcoming
exam.
15. I am willing to allow a student with a documented 102030405060
disability to complete "extra credit" assignments
for academic success even when this option is not
stated on the course syllabus.
16. I am willing to allow any student to complete 102030405060
"extra credit" assignments in my course(s).
17. I am willing to reduce the overall course reading 102030405060
load for a student with a documented disability
even when I would not allow a reduced reading
load for other students.
18. I make individual accommodations for students 10 20 30 40 50 60
who have disclosed their disability to me.
19. I believe that my overall teaching style permits all 102030405060
students to learn the course material regardless of
their individual needs.
92
20. I am willing to allow students with documented 102030 40 50 60
disabilities to record class sessions when necessary.
21. I am willing to extend the "due dates" of 102030 405060
assignments to accommodate the needs of students
with documented disabilities when necessary.
22. Providing teaching accommodations to students 10 20 3D 40 50 60
with documented disabilities is unfair to students
without disabilities.
23. When students with disabilities are having 102030 405060
difficulties in my course(s), I am uncertain about
where I can find additional support at the
University of Oregon.
24. I believe that students with disabilities use the 10 20 3D 40 50 60
disability as an excuse when they are not doing
well in my class.
25. I am willing to arrange extended time on exams for 102030 40 50 60
students who have documented disabilities.
26. I am willing to change the method of responding on 102030 40 50 60
exams (e.g., from written to oral) for students with
documented disabilities.
27. I am willing to spend extra time (i.e., in addition to 102030 405060
normal office hours) helping any student prepare
for an exam or review course material.
28. I feel very comfortable meeting with students to 10 20 3D 40 50 60
discuss their disability-related accommodation
needs.
93
29. I am willing to allow students with documented 102030405060
disabilities to take proctored exams in a supervised
location outside of the normal exam location.
30. I am willing to allow students with documented 102030405060
disabilities to use technology (e.g., laptop,
calculator, spell checker) to complete tests even
when such technologies are not permitted for use
by students without disabilities.
31. Providing testing accommodations (such as 102030405060
extended exam time) to students with documented
disabilities is unfair to students without disabilities.
32. I feel very comfortable discussing with students 102030405060
barriers they may experience given the design or
evaluation methods used in my course(s).
33. Typically, students with disabilities do not perform 102030405060
as well as the rest of the students in my course(s).
34. I am aware of assistive technology that students 102030405060
with disabilities can use to aid their understanding
of course material.
35. Disability Services will provide exam proctoring 102030405060
for students with documented disabilities.
36. I incorporate disability-related topics into the 102030405060
content of my course(s).
37. I routinely build in some flexibility in my 102030405060
evaluation methods, such as dropping the lowest
quiz score or providing different exam formats
(essay, multiple choice, papers) for all students.
38. Students with documented disabilities who request 102030405060
support from Disability Services receive adequate
services from that office.
94
39. I receive adequate support from Disability Services 102030405060
to make appropriate accommodations for students
with documented disabilities.
40. Currently, I do not have sufficient knowledge to 102030405060
make adequate accommodations for students with
disabilities in my course(s).
41. Making adequate accommodations for students 102030405060
with documented disabilities in my courses is
umealistic given time constraints and other job
demands.
42. As a faculty member, I am legally required to 102030405060
provide requested accommodations to students with
disabilities.
43. I have provided recommended accommodations 102030405060
(e.g. extended time, private testing room, etc.) for
students with learning disabilities in my course(s)
44. I have provided recommended accommodations 102030405060
(e.g. large-print materials, captioned videos, etc.)
for students with disabilities in my course(s).
45. When students approach me with accommodation 102030405060
requests they typically show me a letter that lists
recommended accommodations.
46. I include a statement in my syllabus inviting 102030405060
students with disabilities to discuss their needs with
me.
47. I make a statement in class inviting students with 102030405060
disabilities to discuss their needs with me.
95
48. If students with documented disabilities wait to talk 102030405060
to me until they are not doing well in my class then
I feel it's too late to provide appropriate
accommodations.
49. Typically, I use scaffolding methods to assist 10203040 5060
students with assignments in my course(s), such as
using past assignments as examples or requiring
students submit an outline and/or annotated
bibliography weeks before submitting a final paper.
50. I put my lecture notes online for all students (on 102030405060
Blackboard or another website).
51. I would like to modify curricular aspects of my 102030405060
course(s) in order to create a more inclusive
learning environn1ent for all students.
52. I prefer to use a variety of instructional formats in 1020 30 40 50 60
my class, including small group and hands on
activities.
53. I believe it is my responsibility as an instructor to 10 20 30 40 50 60
provide consistent information and expectations on
all assignments and exams.
54. At times, I feel burdened when students with 102030405060
disabilities approach me with accommodation
requests.
55. I refer students to academic support services on 102030405060
campus, such as Academic Learning Services
(ALS), when they need extra help.
56. I am willing to use technology so that my course 10203040 5060
material can be available in a variety of formats
(e.g. podcast of lecture available for download,
course readings available as mp3 files).
57. Students with disabilities should be able to perform 102030405060
just as well as students without disabilities in my
course(s).
96
APPENDIXB
FIRST RECRUITMENT EMAIL TO FACULTY PARTICIPANTS
Dear Faculty Member,
Recently, the UO received grant funding from the Office of Postsecondary Education to
implement Project ExCEL- UO, Expanding Cultural Awareness of Exceptional Learners
at the University of Oregon. The goal of Project ExCEL is to improve the experience of
students with disabilities at the UO by providing all faculty members with additional
information and training related to understanding and teaching students with disabilities.
This professional development model will be implemented collaboratively by faculty in
Special Education and Clinical Sciences and Disability Services. The model includes
three interrelated training components that together are designed to impact the overall
culture ofthe university in ways that make it more responsive to the needs of students
with disabilities.
In order to effectively use the grant funding, a survey has been designed for all teaching
faculty. If you are currently teaching, or have taught classes in the past, please take the
time to complete the survey. Your input is valuable; it will help us assess the current
campus climate and plan future instructional opportunities for the duration of the grant.
The actual survey should take about 20 minutes to complete. In addition, please accept an
Allann Brothers discount coupon as a token of our appreciation for your time. Both the
survey and the coupon are available at the link:
http://www.uoregon.edu/~allisonl/
There is a written consent form at the beginning of the survey. By indicating "yes" you
are giving consent to participate. Your participation is voluntary. Thank you in advance
for your time and input. If you have any question, please contact any of the project staff
listed below.
The Project ExCEL team
Dr. Christopher Murray, Principal Investigator
346-1445; cjmurray@uoregon.edu
Dr. Hilary Gerdes, Director of Disability Services
346-1063; hgerdes@uoregon.edu
Allison Lombardi, Proj ect Coordinator
255-9405; allisonl@uoregon.edu
Leslie Gilbert, Project Assistant
346-3173; 19i1bert@uoregon.edu
97
98
APPENDIXC
LETTER OF CONSENT TO PARTICIPATE: FACULTY
You are invited to participate in a research study conducted by Dr. Christopher Murray, a
faculty member, and Allison Lombardi, a doctoral student, from the Special Education
and Clinical Sciences Department, as well as Hilary Gerdes, Director ofDisability
Services, at the University of Oregon. We hope to learn about faculty attitudes and
perceptions toward disability in order to implement a professional development model
that enhances disability awareness. In addition, Ms. Lombardi will conduct a validity
study on the survey instrument as part of a dissertation study. You were selected as a
possible participant in this study because you are a member of the teaching faculty at the
University of Oregon.
Ifyou decide to participate, please indicate "yes" to the question below to begin the
survey. The survey should take about 20 minutes to complete. Your responses will be
completely anonymous to the researchers. Your participation is voluntary.
The survey questions will ask about your attitudes and perceptions toward students with
disabilities as an instructor. We cannot guarantee that you personally will receive any
benefits from this research. In addition, each participant will be offered a discount
coupon to Allann Brothers.
If you have any questions about the research study, please contact a member of the
Project Excel team:
Dr. Christopher Murray, Principal Investigator
346-1445
£i!purray@uoregon.edu
Dr. Hilary Gerdes, Director of Disability Services
346-1063
hgerdes~uoregon.edu
Allison Lombardi, Project Coordinator
255-9405
allisonl@uoregon.edu
Do you agree to participate in the study?
o Yes
o No
99
100
APPENDIXD
SECOJ\ID EMAIL TO FACULTY PARnCIPANTS
Dear UO Faculty Member:
This email is a reminder to all teaching faculty to please complete the Project ExCEL
Faculty Survey at your earliest convenience. If you have already completed the survey,
thank you very much for your input and please disregard this email. If you have not yet
completed the survey, it should take about 20 minutes to complete. We realize you are
very busy and we appreciate your time. Your input is invaluable to us and the success of
our project. In addition, please accept an Allann Brothers discount coupon as a token of
our appreciation for your time. Both the survey and the coupon are available at the link:
http://www.uoregon.edu/~allisonl/
There is a written consent form at the beginning of the survey. By indicating "yes" you
are giving consent to participate. Your participation is voluntary.
The Project ExCEL team
Dr. Christopher Murray, Principal Investigator
346-1445
cjmurray@uoregon.edu
Dr. Hilary Gerdes, Director of Disability Services
346-1063
hgerdes@uoregon.edu
Allison Lombardi, Project Coordinator
255-9405
allisonl@uoregon.edu
101
APPENDIXE
THIRD EMAIL TO FACULTY PARTICIPANTS
Dear UO Faculty Member:
This is another friendly reminder to please complete the Project ExCEL
Faculty Survey at your earliest convenience. If you have already completed
the survey, thank you very much for your input and please disregard this
email. If you have not yet completed the survey, it should take about 20
minutes to complete. Unfortunately, our response rate has been lower than
anticipated so far and your input is crucial to the success of our project.
We realize you are very busy, but we do hope that you could find some time
to complete the survey. Thank you again! The link is:
http://www.uoregon.edu/~allisonl/
There is a written consent form at the beginning of the survey. By indicating "yes" you
are giving consent to participate. Your participation is voluntary.
The Project ExCEL team
Dr. Christopher Murray, Principal Investigator
346-1445
cjmurrayra1uoregon.edu
Dr. Hilary Gerdes, Director of Disability Services
346-1063
hgerdes@uoregon.edu
Allison Lombardi, Project Coordinator
255-9405
allisonl@uoregon.edu
----~------
102
APPE1\TDIX F
FOlJRTH EMAIL TO FACULTY PARTICIPANTS
Dear UO Faculty Member:
This is the final reminder to all teaching faculty to please complete the Project ExCEL
Faculty Survey at your earliest convenience. If you have already completed
the survey, thank you very much for your input and please disregard this
email. We realize you are very busy, but we do hope that you could find some time to
complete the survey. Your input will greatly impact the success of our proj ect and help
UO students with disabilities. Also, we offer a discount coffee coupon to all teaching
faculty as a token of our appreciation for your time. You will find both the survey and
coupon at the link below:
http://www.uoregon.edu/~allisonl/
There is a written consent form at the beginning of the survey. By indicating "yes" you
are giving consent to participate. Your participation is voluntary.
The Project ExCEL team
Dr. Christopher Murray, Principal Investigator
346-1445
cjmurray@uoregon.edu
Dr. Hilary Gerdes, Director ofDisability Services
346-1063
hgerdes@uoregon.edu
Allison Lombardi, Project Coordinator
255-9405
allisonl@uoregon.edu
103
APPENDIXG
SECURITY QUESTIONS
Please answer the following security questions below prior to beginning the Project Excel
faculty survey. Your answers to these questions will generate a unique code in our
system, which will allow your responses to be confidential.
1. What are the first two letters of your mother's maiden name?
---
2. What is your birth month? (If single digit, place a 0 in front, such as 01 for
January) __
3. What are the last two letters ofthe city where you were born? _
4. What are the last two digits ofthe year you were born? _
APPENDIXH
ITEM MEANS AND STANDARD DEVIATIONS
Item M SD N
Q1: I am confident in my understanding of section 504 2.57 1.46 95
of the Rehabilitation Act of 1973
Q2: I am confident in my understanding of the legal 3.28 1.49 95
definition of disability
Q3: I am confident in my understanding of the 3.34 1.50 95
Americans with Disabilities Act (1990)
QlO: I am willing to provide copies of my lecture notes 5.16 0.93 95
or outlines to students with documented disabilities
Q12: I am willing to provide copies of my overheads 5.43 0.73 95
and/or PowerPoint presentations to students with
documented disabilities
Q13: Providing acconmlOdations to students with 5.33 0.74 95
disabilities is a way to ensure equal opportunity and
access to learning in higher education settings
Q14:1 am willing to spend extra time (i.e., in addition to 5.10 0.92 95
typical office hours) meeting with students with
documented disabilities to clarify and/or review course
related content or prepare for an upcoming exam
Q15: I am willing to allow a student with a documented 3.45 1.52 95
disability to complete extra credit assignments for
academic success even when this option is not stated on
the course syllabus
Q 16:1 am willing to allow any student to complete extra 3.02 1.52 95
credit assignments in my course(s)
Q22:Providing teaching accommodations to students 1.94 0.79 95
with documented disabilities is unfair to students
without disabilities
Q24: I believe that students with disabilities use the 2.01 0.83 95
disability as an excuse when they are not doing well in
my class
Q25: I am willing to arrange extended time on exams 5.48 0.66 95
for students who have documented disabilities
Q27: I am willing to spend extra time (i.e., in addition 4.57 1.29 95
to normal office hours) helping any student prepare for
an exam or review course material
104
105
Item M SD N
Q28: I feel very comfortable meeting with students to 5.26 0.77 95
discuss their disability-related accommodation needs**
Q29: I am willing to allow students with documented 5.55 0.58 95
disabilities to take proctored exams in a supervised
location outside of the normal exam location
Q32: I feel very comfortable discussing with students 4.96 0.91 95
barriers they may experience given the design or
evaluation methods used in my course(s)**
Q37: I routinely build in some flexibility in my 4.17 1.50 95
evaluation methods, such as dropping the lowest quiz
score or providing different exam formats (essay,
multiple choice, papers) for all students**
Q38: Students with documented disabilities who request 4.39 0.78 95
support from Disability Services receive adequate
services from that office
Q39: I receive adequate support from Disability 4.35 0.97 95
Services to make appropriate accommodations for
students with documented disabilities
Q4l: Making adequate accommodations for students 2.18 0.92 95
with documented disabilities in my courses is
umealistic given time constraints and other job demands
Q46: I include a statement in my syllabus inviting 4.30 1.72 95
students with disabilities to discuss their needs with me
Q47: I make a statement in class inviting students with 3.91 1.60 95
disabilities to discuss their needs with me
Q5l: I would like to modify curricular aspects of my 4.18 1.22 95
course(s) in order to create a more inclusive learning
environment for all students**
Q52: I prefer to use a variety of instructional formats in 4.92 1.22 95
my class, including small group and hands on
activities**
Q53: I believe it is my responsibility as an instructor to 5.44 0.66 95
provide consistent information and expectations on all
assignments and exams**
Q55: I refer students to academic support services on 4.84 1.06 95
campus, such as Academic Learning Services (ALS),
when they need extra help**
Q57: Students with disabilities should be able to 4.87 0.96 95
perform just as well as students without disabilities in
my course(s)
APPENDIX I
ITEM CORRELATION MATRIX
106
Item Q1 Q2 Q3 Q\O Q12 Q13 Q14 Q15 Q16 Q22 Q24 Q25 Q27 Q28 Q29 Q32 Q37 Q38 Q39 Q41 Q46 Q47 Q51 Q52 Q53 Q55 Q57
Q1 1.000
Q2 0.729 1.000
Q3 0,752 0.825 1.000
QIO 0.081 0.082 0.082 1.000
Q12 0078 0.Q38 0.090 0646 1.000
Q13 0.015 0.172 0.141 0.137 0.214 1.000
Q14 0.241 0.178 0.154 0.156 0.160 0.236 1.000
Q15 0097 0058 -0.006 0.029 0.058 -0.047 0.213 1.000
Q16 0.194 0.122 0.079 0.144 0.138 -0.005 0.099 0.622 1.000
Q22 -0.168 -0.168 -0.134 -0.190 -0.180 -0.462 -0.176 -0.184 -0.047 1.000
Q24 -0.142 -0.101 -0. [66 -0.043 -0.075 -0.365 -0.360 -0071 0.\00 0.547 1.000
Q25 0.019 0.024 0.072 0.107 0.182 0.385 0.329 0.041 -0089 -0.210 -0.299 1.000
Q27 0.2[7 0.240 0.170 0.257 0.239 0.145 0.747 0.132 0.133 -0050 -0.157 0.240 1.000
Q28 0.232 0.199 0.160 0.131 0.4\4 0.385 0.302 0.008 0.027 -0.250 -0.269 0.353 0293 1.000
Q29 0.204 0.162 0229 0.112 0.177 0.265 0.226 -0125 -0.124 -0.139 -0.272 0.56[ 0.171 0.242 1.000
Q32 0.187 0229 0.089 0.067 0.180 0.272 0.342 0.219 0.215 -0.299 -0.310 0160 0.290 0.484 0167 1.000
Q37 0.102 0.070 -0.042 0.164 0.148 0.022 0075 0100 0.239 -0.093 0.002 0.024 0.188 0.156 0.042 0.186 1.000
Q38 0.003 0.112 0.075 -0.052 ·0.101 0.162 0.073 -0.058 0.063 -0.015 -0062 0.032 0037 -0.011 0.027 -0.032 0053 1.000
Q39 0.037 0.074 0.109 -0.001 -0.03 [ 0.128 0.100 -0.156 -0.015 -0.100 -0.125 0.153 O. [36 0.082 0.167 -0.032 -0023 0.676 1.000
........
0
-.....l
Item QI Q2 Q3 QIO Q12 Q13 QI4 QI5 Q16 Q22 Q24 Q25 Q27 Q28 Q29 Q32 Q37 Q38 Q39 Q41 Q46 Q47 Q51 Q52 Q53 Q55 Q57
Q41 -0.279 -0.252 -0.302 -0.224 -0.310 -0.432 -0.353 -0.065 -0.112 0.402 0.589 -0.375 -0.363 -0.297 -0.348 -0.392 -0029 -0.136 -0.218 1.000
Q46 0.202 0329 0.283 0.068 0.022 0.002 0.187 0.136 0.175 -0.058 -0.020 0.175 0.105 0.135 0.009 0.011 -0.021 0.129 0.076 -0034 1.000
Q47 0.323 0.337 0.302 0.084 0.153 0.088 0.207 0.034 0.081 -0.084 -0.024 0.1 I 1 0.188 0.268 0.128 0.148 0.099 0.033 -0.061 -0.040 0.592 1.000
Q51 0.175 0.147 0.183 0.377 0.277 0.041 0.211 0.269 0.255 -0.244 0026 0.179 0.210 0.155 0.148 0.154 0.109 0.045 0.048 -0.107 0.412 0.278 1.000
Q52 0.234 0.263 0.236 0.230 0.260 0.116 0.236 0.040 0.147 -0.001 -0.004 -0.027 0.281 0.114 0.082 0.047 0.068 0.050 -0.019 -0.076 0.252 0.221 0.514 1.000
Q53 0.137 0.139 0.106 0.106 0.080 0.210 0.349 -0.063 0.023 -0.187 -0.269 0.299 0.272 0.332 0.254 0.200 0.150 0.110 0.074 -0.184 0.293 0.266 0.244 0.267 1.000
Q55 0.094 0.181 0.225 -0.001 0.035 0.186 0.056 0.013 -0.054 -0.204 -0.233 0.159 0.028 0.266 0.269 0.051 0.021 0.094 0.126 -0.073 0.099 0.252 0.209 0.01\ 0243 1.000
Q57 0.168 0152 0.200 0.162 0.142 0.250 0.247 -0.126 0.113 -0.261 -0.221 0.165 0.204 0.340 0.181 0.160 -0.017 0.230 0.307 -0.182 0.294 0.221 0.297 0.268 0.390 0.167 1.000
>---'
o
00
109
REFERENCES
American Educational Research Association, American Psychological Association, &
National Council on Measurement in Education. (2009). Standards for
Educational and Psychological Testing. Washington, DC: American Educational
Research Association.
Arbuckle, 1. L. (2006). AMOS 7. O. Amos Development Corporation.
Askamit, D., Morris, M., & Luenberger, 1. (1987). Preparation of student services
professionals and faculty for serving learning-disabled college students. Journal
ofCollege Student Personnel, 28(3),53-59.
Beilke, 1. R & Yssel, N. (1999). The chilly climate for students with disabilities in higher
education. College Student Journal, 33(3), 364-372.
Blacqua, T. c., Rapaport, R J., & Kruse, B. G. (1996). Perceptions ofclassroom
accommodations among college students with disabilities. ERIC Document
Reproduction Service No. ED 400 640.
Burgstahler, S. (2008). Universal design ofinstruction: Definition, principles, and
examples. Retrieved June 16,2008, from
http://www.washington.edu/doitlBrochures/Academicslinstruction.html
Burgstahler, S., Duclos, R, & Turcotte, M. (1999). Preliminary findings: Faculty,
teaching assistant and student perceptions regarding accommodating students
with disabilities in postsecondary environments. Seattle, WA: University of
Washington, DO-IT.
Chubon, R. A (1992). Attitudes toward disability: Addressing fundamentals of attitude
theory and research in rehabilitation education. Rehabilitation Education, 6, 301-
312.
Clark, L., & Watson, D. (1995). Constructing validity: Basic issues in objective scale
development. Psychological Assessment, 7(3),309-319.
Cohen, J. (1973). Eta-squared and partial eta-squared in fixed factor ANOVA designs.
Educational and Psychological Measurement, 33, 107-112.
Connell, B. R, Jones, M., Mace, R, Mueller, 1., Mullick, A, Ostroff, E. (1997). 171e
principles ofuniversal design. Retrieved August 23, 2008, from
http://www.design.ncsu.edu/cud/about_ud/udprincipleshtmlformat.html
110
Cope, D. (2005). Disability law and your classroom. Academe Online, 91(6).
Costello, A. B., & Osborne, J. W. (2005). Best practices in exploratory factor analysis:
Four recommendations for getting the most from your analysis. Practical
Assessment, Research & Evaluation, 10(7). Available online:
http://pareonline.net/pdf/v1On7.pdf
Dempster, A. P., Laird, N. M., & Rubin, D. B. (1977). Maximum likelihood estimation
from incomplete data via the EM algorithm (with discussion). Journal ofthe
Royal Statistical Society, Series B, 39, 1-38.
Deshler, D. D., Ellis, E. S., & Lenz, B. K. (1996). Teaching adolescents with mild
disabilities. Denver, CO: Love Publishing Company.
Dillman, D. A. (2007). Writing questions. In D.A. Dillman (ed.) Mail and Internet
Surveys: The Tailored Design Method (2nd Ed.). (pp. 32-78). New York: John
Wiley.
Dowrick, P. W., Anderson, 1., Heyer, K., & Acosta, J. (2005). Postsecondary education
across the USA: Experiences of adults with disabilities. Journal of Vocational
Rehabilition, 22,41-47.
Fabrigar, L. R., Wegener, D. T., MacCallum, R. c., & Strahan, E. R. (1999). Evaluating
the use of exploratory factor analysis in psychological research. Psychological
Methods, 4(3),272-299.
Farone, M. C., Hall, E. W., & Costello, J. J. (1998). Postsecondary disability issues: An
inclusive identification strategy. Journal ofPostsecondary Education and
Disability, 13, 35-45.
Fonosch, G. G., & Schwab, L. O. (1981). Attitudes of selected university faculty
members toward disabled students. Journal and College Student Personnel, 22,
229-235.
Hartman-Hall, H. M., & Haaga, D. A. (2002). College students' willingness to seek help
for their learning disability. Learning Disability Quarterly, 25, 263-274.
Haynes, S. N., Richard, D. C. S, & Kubany, E. S. (1995). Content validity in
psychological assessment: A functional approach to concepts and methods.
Psychological Assessment, 7(3),238-247.
Henson, R. K. (2001). Understanding internal consistency reliability estimates: A
conceptual primer on coefficient alpha. Measurement and Evaluation, 34, 177-
189.
111
Johnson, B., & Stevens, J. J. (2001). Exploratory and confinnatory factor analysis of the
School Level Environment Questionnaire (SLEQ). Learning Environments
Research, 4, 325-344.
Johnson, B., Stevens, J. J., & Zvoch, K. (2007). Teachers' perceptions of school climate:
A validity study of scores from the revised school level environment
questionnaire. Educational and Psychological Measurement, 67(5),833-844.
Ketterlin-Geller, L. & Johnstone, C. (2008). Accommodations and universal design:
Supporting access to assessments in higher education. Journal ofPostsecondary
Education and Disability, 19(2), 163-171.
Kline, R (1998). Principles and practice ofstructural equation modeling. New York:
Guilford Press.
Levine, T. R., & Hullett, C. R (2002). Eta squared, partial eta squared, and misreporting
of effect size in communication research. Human Communication Research,
28(4),612-625.
Madaus, J. W., Scott, S. & McGuire, J. (2003a). Addressing student diversity in the
classroom: The approaches ofoutstanding university professors (Universal
Design for Instruction Project Technical Report No. 02). Retrieved from
University of Connecticut, Center on Postsecondary Education and Disability
website: http://www.facultyware.uconn.edu/TechnicalReports.cfm
Madaus, J. W., Scott, S. & McGuire, J. (2003b). Barriers and bridges to learning as
perceived by postsecondary students with learning disabilities (Universal Design
for Instruction Project Technical Report No. 01). Retrieved from University of
Connecticut, Center on Postsecondary Education and Disability website:
http://www.facultyware.uconn.edu/TechnicalReports.cfm
Matthews, P. R, Anderson, D. W., & Skolnick, B. D. (1987). Faculty attitudes toward
accommodations for college students with learning disabilities. Learning
Disabilities Focus, 3, 46-52.
Messick, S. (1980). Test validity and the ethics of assessment. American Psychologist,
35(11),1012-1027.
Messick, S. (1995). Validity of psychological assessment. American Psychologist, 50(9),
741-749.
Mull, c., Sitlington, P. L., & Alper, S. (2001). Postsecondary education for students with
learning disabilities: A synthesis of the literature. Exceptional Children, 97-118.
112
Murray, C., Goldstein, D. E., Nourse, S. & Edgar, E. (2000). The postsecondary school
attendance and completion rates of high school graduates with learning
disabilities. Learning Disabilities Research & Practice, 15, 119-127.
Murray, C., Lombardi, A., Wren, C., & Keys, C. (2009). The effects of disability-focused
training on attitudes and perceptions of university staff. Learning Disability
Quarterly, 32(2),87-102.
Murray, C., Wren, c., & Keys, C. (2008). University faculty perceptions of students with
learning disabilities: Correlates and group differences. Learning Disability
Quarterly, 31, 1-19.
Murray, C., Wren, c., Stevens. E., & Keys, C. (2009). Promoting university faculty and
staff awareness of students with learning disabilities: An overview of the
productive learning u strategies (PLUS) project. Journal ofPostsecondary
Education and Disability, 22(2), 117-129.
Nelson, J. R, Dodd, J. M., & Smith, D. J. (1990). Faculty willingness to accommodate
students with learning disabilities: A comparison among academic divisions.
Journal ofLearning Disabilities, 23(3), 185-189.
Nunnally, J. C. (1975). Psychometric theory: 25 years ago and now. Educational
Researcher, 4(10), 7-21.
O'Connor, RP. (2000). SPSS and SAS programs for determining the number of
components using parallel analysis and Velicer's MAP test. Behavior Research
Methods, Instruments, and Computers, 32(3),396-402.
Ouellett, M. L. (2004). Faculty development and universal instructional design. Equity
and Excellence in Education, 37, 135-144.
Parker, D. R, Embry, P. R, Scott, S. S., & McGuire, J. M. (2003). Disability service
providers I perceptions about implementing Universal Design for Instruction
(UDI) on college campuses (Universal Design for Instruction Project Technical
Report No. 03). Retrieved from University of Connecticut, Center on
Postsecondary Education and Disability website:
http://www.facultyware.uconn.edu/TechnicaIReports.cfin
Patton, M. Q. (1982). Thoughtful questionnaires. In M.Q. Patton (ed.) Practical
Evaluation. (pp. 139-159). Newbury Park, CA: SAGE.
Preacher, K. P., & MacCullum, R C. (2003). Repairing tom swift's electric factor
analysis machine. Understanding Statistics, 2(1), 13-43.
113
Rao, S. (2004). Faculty attitudes and students with learning disabilities in higher
education: A literature review. College Student Journal, 38(2), 191-198.
Reed, M. J., Lund-Lucas, E., & O'Rourke, K. (2003). Standards of practice in
postsecondary special needs programming: Student and administrator opinion.
The Canadian Journal ofHigher Education, 2, 27-56.
Scott, S. S., McGuire, J. M., & Shaw, S. F. (2003). Universal design of instruction: A
new paradigm for adult instruction in postsecondary education. Remedial and
Special Education, 24(6),369-379.
Shaw, S. F., & Dukes, L .L. (2001). Program standards for disability in higher education.
Journal ofPostsecondary Education and Disability, 14(2), 81-90.
Skinner, M. E. (2007). Faculty willingness to provide accommodations and course
alternatives to postsecondary students with learning disabilities. International
Journal ofSpecial Education, 22(2), 32-45.
SPSS, Inc. (2006). SPSS 15.0 for Windows Graduate Student Version. Lead
Technologies.
Stevens, J. P. (2002). Applied Multivariate Statistics for the Social Sciences. (4th ed.)
Mahweh, NJ: Lawrence Erlbaum Associates, Publishers.
Stevens, J. J., & Zvoch, K. (2007). Confirmatory factor analysis and the terra nova
comprehensive tests of basic skills. Educational and Psychological Measurement,
67(6), 976-989.
Tabachnick, B. G., Fidell, L. S. (2001). Using multivariate statistics (4th Ed.). New York:
Harper & Row.
U.S. Department ofEducation, National Center for Education Statistics. (1999). Students
with disabilities in postsecondary education: A profile ofpreparation,
participation, and outcomes. NCES 1999-187. Washington, DC: U.S.
Government Printing Office.
U.S. Department of Health and Human Services. (2003). Trends in the well-being of
America's children and youth. Washington. DC: U.S. Government Printing
Office.
U.S. Department of Labor (2007). College enrollment and work activity of2006 high
school graduates (USDL 07-0604). Retrieved March 12,2008
http://www.bls.gov/news.release/hsgec.nrO.htm.
114
U.S. Department of Labor (2008). Education pays in higher earnings and lower
unemployment rates. Retrieved May 1, 2008,
http://www.bls.gov/emp/emptab7.htm
University of Oregon, Office of Affirmative Action and Equal Opportunity. (2007).
Affirmative Action Plan. Retrieved October 27,2008 from:
http://aaeo.uoregon.edu/plan.html
University of Oregon, Office ofInstitutional Research. (2007). Faculty counts by type.
Retrieved October 23,2008 from: http://ir.uoregon.edu/data
Vogel, S. A., Leyser, Y, Wyland, S., & Brulle, A. (1999). Students with learning
disabilities in higher education: Faculty attitudes and practices. Learning
Disabilities Research & Practice, 14, 173-186.
Vogel, S. A., Holt, J. K., Sligar, S., & Leake, E. (2008). Assessment of campus climate to
enhance student success. Journal ofPostsecondary Education and Disability,
21(1),15-31.
Wagner, M., Newman, L., Cameto, R., Levine, P, and Garza, N. (2006). An Overview of
Findings From Wave 2 ofthe National Longitudinal Transition Study-2 (NLTS2).
Menlo Park, CA: SRI International. Retrieved October 20, 2007 from:
www.nlts2.org/reports/2006_08/
Wagner, M., Newman, L., Cameto, R., Garza, N., and Levine, P. (2005). After High
School: A First Look at the Postschool Experiences of Youth with Disabilities. A
• Report from the National Longitudinal Transition Study-2 (NLTS2) Menlo Park,
CA: SRI International.
Wagner, M., Newman, L., Cameto, R. (2004). Changes over time in the secondary school
experiences ofstudents with disabilities. A report offindings from the National
Longitudinal Transition Study (NLTS) and the National Longitudinal Transition
Study-2 (NLTS-2). Menlo Park, CA: SRI International.
Yuker, H. E., Block, J. R., & Campbell, W. J. (1960). A scale to measure attitudes
toward disabled persons. Albertson, NY: Human Resources Center.