Item 08 - Ballot Measure Polling Results

Home Browse Search

M E T H O D O L O G Y The following sections outline the methodology used in the study, as well as the motivation for using certain techniques. 0I11:STIONNAIRE DEVELOPMENT Dr. McLarney of True North Research worked closely with the City of Santa Ana to develop a questionnaire that covered the topics of interest and avoided possible sources of systematic measurement error, including position -order effects, wording effects, response -category effects, scaling effects, and priming. Several questions included multiple individual items. Because asking the items in a set order can lead to a system- atic position bias in responses, items were asked in random order for each respondent. Some questions asked in this study were presented only to a subset of respondents. For exam- ple, only individuals who did not support the sales tax measure (or were unsure) at the Initial Bal- lot Test (Question 5) were asked Question 6 regarding their reasons for not supporting the measure. In some cases, two versions of a project or argument were tested to identify how word- ing or framing differences impact perception of the item. In such cases, half of the sample received the item with version 1 wording (e.g., Question 8, item C1) and the other half received version 2 (e.g., Question 8, item C2). The questionnaire included with this report (see Question- naire & Toplines on page 31) identifies the skip instructions that were used during the interview to ensure that each respondent received the appropriate questions. IrKU(,K/AiviiVIINC,, F'KL- I LS I & I KANSLA I IUN Prior to fielding the survey, the ques- tionnaire was CATI (Computer Assisted Telephone Interviewing) programmed to assist interview- ers when conducting telephone interviews. The CATI program automatically navigates skip patterns, randomizes the appropriate question items, and alerts the interviewer to keypunching mistakes should they occur. The survey was also programmed into a passcode-protected online survey application to allow online participation for sampled voters. The integrity of the question- naire was pre -tested internally by True North and by dialing into voter households in the City prior to formally beginning the survey. Once finalized, the questionnaire was professionally translated into Spanish and Vietnamese to allow for data collection in three languages. SAMPLE The survey was administered to a stratified and clustered random sample of regis- tered voters in the City who are likely to participate in the November 2026 general election, with a subset who are also likely to participate in the lower -turnout June 2026 primary. Consistent with the profile of this universe, the sample was stratified into clusters, each representing a com- bination of age, gender, and household party type. Individuals were then randomly selected based on their profile into an appropriate cluster. This method ensures that if a person of a par- ticular profile refuses to participate, they are replaced by an individual who shares their same profile. PT A -1-1 - A - - - - "`-'"' "r- r- -- -- By using the probability -based sampling design noted above, True North ensured that the final sample was representative of voters in the city who are likely to participate in the November 2026 general election. The results of the survey can thus be used to estimate the opinions of all voters likely to participate in said election. Because not all voters participated in the study, however, the results have what is known as a sta- tistical margin of error due to sampling. The margin of error refers to the difference between City of Santa Ana True North Research, Inc. © 2025