A new opinion poll was released this week by Noble Predictive Insights, regarding the relative standing of the two main contenders for the Republican nomination for governor at the 2026 primary.
Those of us who have watched political polling for some time have learned that polls have some unique characteristics, among them:
- They may not be totally accurate. Even polls conducted by experienced professionals, often fall short in the accuracy department.
- They cannot please everyone. Those who agree with the results of the polls think they are great. Those who do not agree, immediately question their validity and some times even the motives and integrity of those conducting the polls.
- Polls are only snapshots of opinions at one point in time. The same poll taken weeks or even days before or after may show a different outcome.
- Different pollsters may use different methodologies, resulting in different results.
- Polls should not be taken too seriously. They are useful tools to help policy makers make decisions, but should not be the only, or even the main, tool to be used.
This latest poll is causing more than the usual level of controversy for three reasons:
- The results are contrary to the common perception as to who will be, or should be, the primary election winner.
- Recent election irregularities have made many voters quite skeptical of any information provided by anyone considered to be part of the establishment.
- A simple typographical error on the results graphic awakened feelings in some folks that this was not a real poll.
This last point warrants further explanation. When the original email was sent out, the footnote on the graphic stated that the survey was conducted August 11-18, 2024. Some of those who disagreed with the result wasted no time leveling accusations of all kinds of foul play.
Their pre-conceived ideas about the pollster prevented them from considering that this may have been the result of a simple typo.
They did not bother to read the methodology section, where the date was correctly stated as August 11-18, 2025.
They did not bother to check with the pollster, as this writer did, to confirm that there was in fact a typographical error.
The original email cannot be recalled, but the correction was made where possible, and the corrected version is the one we are including here.
Methodology: This poll was conducted as an online opt-in panel survey. The survey was completed by Noble Predictive Insights from August 11–18, 2025 from an Arizona statewide registered voter sample. The sample demographics were weighted to accurately reflect gender, region, age, party affiliation, ethnicity, and education according to recent voter file data, the Arizona Secretary of State’s Office, and recent Census data. The sample size was 948 registered voters, yielding a MoE of ± 3.18%. The total sample included 385 registered Republicans, yielding a margin of error of ± 4.99%. Numbers may not equal 100% due to rounding.
Similar surveys have been taken in May and February of this year, and the results, as expected, are different:
May, 2025
| Robson | Biggs | Kirk | Yee | Undecided |
| 24% | 17% | 17% | 6% | 37% |
February, 2025
| Biggs | Kirk | Robson | Jack McCain | Yee | Hoffman | Undecided |
| 14% | 14% | 11% | 8% | 5% | 1% | 47% |
Since future polls, taken by this pollster or others, will likely show different outcomes, it is a good idea to avoid getting too worked up over them, regardless of one’s agreement or disagreement with the results.