First Amendment Survey

Methodology

Participants

 At the time of the investigation the Wisconsin state population was 5,475,649. To best represent the population a telephone survey methodology was employed. For a population of this size a random sample of 1,067 respondents was required to achieve a scientifically valid sample with 95% confidence that results are accurate within +/- 3 % margin of error. A random sample of 600 respondents was required to achieve a scientifically valid sample with 95% confidence that results are accurate within +/- 4 % margin of error.

Through consultation with Survey Sampling International in Fairfield, Connecticut, a random sample of 7,200 Wisconsin telephone numbers was generated and purchased using Random Digit Distribution (RDD) sampling. The telephone numbers included listed and nonlisted numbers. The estimated need for 7,200 numbers was based on the desired 1,067 responses, an estimate of working telephone connections (55%), the likely incidence of a residence having a respondent age 18 or older (90%), and the estimated completion rate based on length of the survey and sensitivity of the subject matter (30%). After a prescreening process to eliminate disconnected and nonworking connections, 5,082 random telephone numbers were provided by Survey Sampling International. These numbers were delivered in 72 separate replicates of approximately 70 random numbers per replicate. No names or addresses of persons were available.

Instruments

Following consultation with the project’s sponsor and a review of related literature and research, an original telephone introductory script (Appendix A) and survey instrument (Appendix B) were developed for the investigation.

The introductory script was designed for delivery within 30 seconds. The telephone script introduced the interviewer calling from the University of Wisconsin-Stout Survey Center, described the general purpose of the survey, assured the respondent that their participation would be voluntary and confidential, and promised to take only four minutes of their time. Respondents were asked if they were age 18 or older prior to beginning interviews. A series of “Fallback” responses was used if respondents had any questions regarding the survey’s purpose, sponsor, use of findings, how they were selected for participation, or to direct respondents to the principal investigator (Appendix A).

The survey instrument was designed to be delivered in approximately four minutes. Several items were similarly worded to national surveys on the First Amendment. The instrument assessed knowledge of the rights afforded by the First Amendment, general opinions on whether First Amendment protections for speech, press and religion are appropriate, whether respondents felt their personal rights had ever been violated, a series of questions regarding opinions on issues relevant to protections in public schools, and an assortment of currently debated issues on free speech, press, religion and assembly. Demographic items included plans to vote in the 2004 presidential election, typical political affiliation, race/ethicity, age, income, and sex. All item responses were in discrete categories. The instrument was revised multiple times to assure content validity and efficiency in interviewing.

Procedures

Following approval for use of human subjects through the UW-Stout Institutional Review Board, telephone interviewing began on March 16, 2004 and continued through March 31, 2004. Calls were made between the hours of 6:00 and 9:00 P.M on weeknights and 1 and 4 P.M. on one Saturday.

Interviews were conducted by the principal investigator, the project manager, the graduate students enrolled in PSYC 791 Research for Decision Making, and student volunteer interviewers. Training sessions of approximately 30 minutes were conducted to prepare interviewers with regard to ethical and efficient delivery of the telephone script and instrument, techniques for encouraging respondent participation, and how to record the disposition of calls on a Call Sheet. All interviewers were asked to sign statements pledging their adequate preparation for conducting interviews and confidentiality of respondents’ comments.

Each telephone call was recorded by interviewers on a Call Sheet with regard to successful completion of the interview, refusal by respondent, nonworking status of the phone, nonresidential status of the phone, partial responses, inability of respondent to reply due to language or disability, no answer, no adult in the home, or answering machine connection. The Call Sheet data was used to identify those telephone numbers eligible for additional call attempts. All telephone numbers were called at least one time. If previous attempts indicated no answer, no adult, or answering machine connections, additional calls were attempted within the data collection time frame.

esponses to the survey were recorded during interviews on optical scan questionnaires (Appendix C) for efficient conversion to statistics software for analysis.