OREGON SURVEY RESEARCH LABORATORY SCHOOL FUNDING IN TIMBER-DEPENDENT COMMUNITIES SURVEY AUGUST, 2002 METHODOLOGY AND RESULTS PAGE 1 SCHOOL FUNDING IN TIMBER-DEPENDENT COMMUNITIES SURVEY: SURVEY METHODOLOGY AUGUST 2002 OREGON SURVEY RESEARCH LABORATORY 5245 UNIVERSITY OF OREGON EUGENE, OR 97408-5245 TELEPHONE: 541-346-0824 FACSIMILE: 541-346-0388 E-MAIL: OSRL@OREGON.UOREGON.EDU WWW: HTTP://OSRL@UOREGON.EDU By Joel D. Bloom, Research Associate and Patricia A. Gwartney, Ph.D., Director with assistance from Deborah McGeehan, Education Program Assistant INTRODUCTION Douglas County School District #77 - Glendale contracted with the University of Oregon Survey Research Laboratory (OSRL) to conduct the ?School Funding in Timber-Dependent Communities Survey.? This survey?s purpose is to assess Oregon residents? opinions and stereotypes related to timber- dependent communities (TDCs) and people who live in them. Working closely with Glendale Schools Superintendent, Brian Metke, and his representatives, Nancy Spens, Dan Clark and Steve Mountainspring, OSRL planned, pretested and implemented a telephone survey of 408 randomly selected households in Oregon. This report summarizes the survey methodology and results. SURVEY METHODOLOGY This section describes OSRL?s procedures for developing and implementing the telephone survey instrument and the random sample required to conduct this representative study of Oregon. SURVEY INSTRUMENT The survey instrument resulted from intensive meetings between Nancy Spens, Dan Clark, and Steve Mountainspring, representing Dr. Metke, and Patricia Gwartney (OSRL Director and professor, sociology, UO) and Joel Bloom (OSRL Research Associate) who collaborated to identify areas to investigate, to decide key concepts, and to operationalize those concepts into survey questions. To provide points of comparison, a few questions replicate or approximate those from previous surveys on stereotyping. Most questions, however, were original. The instrument included individual demographic information needed both to profile respondents and to assess points of variability in the survey results. OSRL staff pretested individual questions for clarity, accuracy, validity, and variability of response. They pretested the entire instrument for flow, comprehensiveness, length, and factors that affect respondents? cooperation and attention. OSRL Project Director Vikas Gumbhir programmed the survey instrument and sample into OSRL?s computer-aided telephone interviewing (CATI) system, and OSRL staff pretested the CATI version. OREGON SURVEY RESEARCH LABORATORY SCHOOL FUNDING IN TIMBER-DEPENDENT COMMUNITIES SURVEY AUGUST, 2002 METHODOLOGY AND RESULTS PAGE 2 1. All interviews were completely anonymous. The University of Oregon?s Committee for the Protection of Human Subjects approved the study?s research design and survey instrument, as required by federal Opinions and evaluations of stereotypes about TDCs and people who live in them ? for this law to safeguard respondents? rights. (See Section 3 of the bound report for significant portions of the human subjects review materials.) The survey interview began by describing the study and guaranteeing respondents anonymity. Since the survey sample was to be drawn from a statewide sample of adults, no special screening questions were necessary. The interview comprised the following subject areas: 1. Respondents? self-assessed level of awareness about TDCs, and their feelings toward them. 2. Opinions and evaluations of stereotypes about TDCs and people who live in them ? for a portion of the survey respondents were randomly assigned either to a set of questions regarding stereotypes about TDCs or about loggers and millworkers. Several questions included both closed-ended and open-ended responses. 3. Opinions regarding whether respondents would want a son to become a logger or millworker or a daughter to marry one (again with random assignment to one or the other). 4. Opinions on whether loggers and millworkers are responsible for ecological damage and whether those occupations are honorable. 5. Demographic and background data ? age, sex, educational attainment, and household location (county, zip code, urban, rural, farm, as well as both respondent self-assessed and externally- assessed location within a TDC). Additional questions asked whether respondents would like to live in a TDC, whether they have ever lived in one, whether they have friends or relatives who do, and how many years they have lived in Oregon. 6. The survey interview concluded with an open-ended question, ?Is there anything you would like to add?? Section 2 of the three-ring binder documenting this study provides a facsimile of the survey instrument with embedded ?topline? frequency results. SURVEY SAMPLE OSRL?s sampling procedure employs a random-digit-dialing (RDD) algorithm used in conjunction with our CATI system. Sampling is pre-programmed and accomplished without interviewers? intervention. Telephone numbers are generated randomly by the computer and appear automatically on interviewers? computer screens. Interviewers place telephone calls with a computer keystroke, effectively preventing dialing errors. This sampling system avoids biases encountered from telephone books and similar lists. In addition, new and unlisted telephone numbers have an equal chance of being selected as established numbers. In order to achieve 408 representative interviews, OSRL randomly generated 1,906 telephone numbers. Altogether, OSRL interviewers dialed those numbers 11,824 times for this study. Those dial attempts distributed as: OREGON SURVEY RESEARCH LABORATORY SCHOOL FUNDING IN TIMBER-DEPENDENT COMMUNITIES SURVEY AUGUST, 2002 METHODOLOGY AND RESULTS PAGE 3 ? 6% phone slams and refusals; ? 3% completed interviews; ? 5% households eligible for survey but not interviewed (e.g., too busy now and adult not home); ? 9% ineligible (e.g., disconnected, nonworking, nonresidential, and fax/modem); and ? 77% study eligibility or ineligibility could not be ascertained (e.g., busy signal, no answer, and answering machine that does not distinguish a household or business). For the 1,907 telephone numbers randomly generated for the study, the final distribution was: ? 54% ineligible (disconnected, nonworking, nonresidential, fax/modem, etc.); ? 21% completed interviews; ? 19% unknown (telephone consistently busy, never answered, or answering machine, and thus suitability for interviewing could not be ascertained); ? 6% phone slams; ? 2% refusal; ? 2% interviews not achieved (respondent ?too busy? or not home); and ? 2% unable to interview (chronic illness, away for study duration, or language barrier). The final survey response rate was 66% and the refusal rate was 8%1. Section 4 of the bound final report provides a complete sample and response rate report, illustrating the final distribution of dial attempts and sample, as well as response and refusal rates, described above. DATA COLLECTION Interviewer training for this survey took place on August 13, 2002; see Section 3 for summary interviewer instructions. Dr. Metke, Nancy Spens and Dan Clark attended the training session. Only experienced interviewers took part in this study?s data collection. OSRL conducted interviewing from 9:00 a.m. until 9:00 p.m. seven days per week (except Sundays, which start at noon) in order to reach citizens with unusual schedules and to avoid nonresponse biases. The target sample size, n=400, was exceeded August 26th with n=408. Completed interviews averaged 8 minutes. All interviews were conducted in English. In administering the survey instrument, OSRL?s trained interviewers use the CATI system, which enables sampling, interviewing, and data entry interactively and seamlessly. Interviewers use telephone headsets in sound-reduced carrels at computer workstations connected by an NT network. CATI randomly distributes telephone numbers to each workstation. The telephone numbers appear automatically on interviewers? computers and CATI mates them to the pre-programmed survey instruments. Interviewers place telephone calls with a computer keystroke, effectively preventing dialing errors. OSRL automatically strips telephone numbers from the interview data to ensure respondents? anonymity. 1 Response rates are calculated in following manner. Completed interview / (Eligible sample + ((Eligible sample / (Eligible sample + Ineligible sample)) * Sample with unknown status)). Source: Robert M. Groves, Survey Errors and Survey Costs, 1989. OREGON SURVEY RESEARCH LABORATORY SCHOOL FUNDING IN TIMBER-DEPENDENT COMMUNITIES SURVEY AUGUST, 2002 METHODOLOGY AND RESULTS PAGE 4 The programmed survey instrument contains all survey questions, interviewer probes for consistency, and pre-coded answer categories. As respondents answer questions, interviewers enter the data into the CATI data file. Skip logic is programmed into the system, preventing inappropriate or incorrect questions from being asked. CATI eliminates out-of-range responses and wild codes by validating each response interactively and disallowing entry of inappropriate responses. Thus, the CATI system eliminates many routine and error-prone coding and data entry tasks and enables OSRL to maintain the highest quality control standards.