March 6, 2011
Wisconsin Study Methodology
This study comprised a representative sample of 603 Wisconsin Adults Age 18 and over and was conducted from February 27 to March 1 of 2011. The focus of the study was assessments of the current political situation and on policy issues with particular attention paid to the current debate about Governor’s Walker's budget repair bill and other budget options. The survey instrument also included standard demographic measures. This survey conforms in full with the disclosure requirements of the American Association of Public Opinion Research (AAPOR) Code of Professional Ethics and Practice and the AAPOR Transparency Initiative.
The study was funded by the Wisconsin Policy Research Institute and was directed by Ken Goldstein. The field work for this survey was carried out by LHK Partners of, Newtown Square PA, using a dual-frame sample design covering both landline telephone and cell phone numbers, with samples produced by Genesys Sampling System. In total, 462 respondents were reached by landline and 141 were reached by phones.
In the landline component of the study, a sample of landline households in the continental United States was selected by GSS via random digit dialing procedures, in which all landline telephone numbers -whether listed and unlisted -- have an equal probability of selection. In the cell phone component of the study, wireless telephone numbers in Wisconsin also had an equal probability of selection in the sample. Due to porting of numbers, cell phone respondents were screened to confirm residence in the state. In addition, all cell phone interviewers were manually dialed. The survey used first birthday selection for sampling within household. Cell phone interviews were conducted with the person who answered the phone.
Interviews for the study were conducted via computer-assisted telephone interviewing (CATI) by professional interviewers trained in interviewing practices, including techniques designed to achieve the highest possible respondent cooperation. Phone numbers were released for interviewing in replicates by region in Wisconsin to allow for proper sample control. Numbers were called multiple times during alternative times of the day during the field period to insure proper sample management. Up to four call-backs were conducted on a sampled number.
For sampling and non-sampling deviations from known population values, data was weighted to the latest US Census’ American Community Survey estimates for Wisconsin for age, sex, and education. In addition, the data were weighted so that each geographic area would correspond to its proper size proportion for the entire state. An iterative raking weight methodology, which has been demonstrated to achieve greater consistency, was utilized, and weights were examined to confirm that there were no unusually large weights. Below are the target weights by gender, age, education, and region for the study.
- 18 to 24 (12.9 percent)
- 25 to 44 (34.7 percent)
- 45 to 64 (35.6 percent)
- 64 plus (16.9 percent)
- Less than High School - (11 percent)
- Completed High School - (35 percent)
- Some College - (21 percent)
- Associates Degree (9 percent)
- Four Year College Degree - (16 percent)
- Graduate or Professional Degree (8 percent)
- City of Milwaukee (10 percent)
- Milwaukee Media Market with city of Milwaukee (29 percent)
- Madison Media Market (19 percent)
- Green Bay Media Market (19 percent)
- Rest of State (24 percent)
- Male (49 percent)
- Female (51 percent)