About the May 2025 Minnesota Community Survey
Conducted May 9 to Aug. 30, 2025, online in English or Spanish
- Who answered
- 1,424
- Minnesota adults
- Completion rate
- 48%
- of 2,990 invited panelists finished
- Margin of error
- ±3.47
- percentage points for all adults; wider for groups
Who responded
Every respondent was first recruited by mail in 2024 from a random sample of Minnesota addresses (a probability-based sample). Any adult in the household could complete the recruitment survey. Recruitment was in English, Spanish, Hmong, and Somali.
Every panelist who agreed to take future surveys was invited by email, postcard, and text message, with two reminders by email, and was offered $5 for completing it. We weighted the results to match Minnesota adults by age, gender, race and ethnicity, education, metro area, volunteering, and 2020 presidential vote.
How the answers were checked
We screened every interview for quality, including a check for respondents who sped through the survey.
How precise the results are
The ±3.47-point margin for all 1,424 respondents already accounts for weighting. Groups such as young adults or Black Minnesotans have fewer respondents, so their margins are wider. We do not report a group with fewer than 50 respondents. Some questions were asked of fewer people; each chart shows its own count.
This survey was paid for by Lumaris Research.
We follow the transparency standards of the American Association for Public Opinion Research. Our published methods and research on survey quality are at lumaris.com/methods.
Technical details
- Unweighted n
- 1,424
- Invited
- 2,990
- Design effect
- 1.78
- Effective sample size
- 800
- Margin of sampling error
- ±3.47 pts, 95% confidence
- Recruitment response rate
- 9% (2024)
Effective sample size reflects precision after weighting. The actual number of respondents is shown as “Unweighted n.”
Weighting
Weighting adjusts how much each respondent’s answers contribute to the results. Two steps: first an adjustment for differing probabilities of selection at recruitment, then calibration to population benchmarks on the dimensions below.
| Dimension | Benchmark source |
|---|---|
| Core demographics | |
| Age (5 groups) | 2022 ACS |
| Gender | 2022 ACS |
| Race/ethnicity | 2022 ACS |
| Education | 2022 ACS |
| Combinations of two demographics | |
| Gender by age (3 groups) | 2022 ACS |
| Education (3 groups) by age (3 groups) | 2022 ACS |
| Education (3 groups) by gender | 2022 ACS |
| Race/ethnicity by education (3 groups) | 2022 ACS |
| Additional | |
| Metropolitan status | 2022 ACS |
| Volunteer status | 2021 CPS Volunteer Supplement |
| 2020 presidential vote | 2020 election results |
ACS: American Community Survey. CPS: Current Population Survey.
Variance
The margins of error use a design-effect approximation based on variation in the final survey weights. The design effect is approximated as 1 plus the squared coefficient of variation of the weights. Sampling error is one source of error; question wording and response accuracy can also contribute. People who did not participate or could not be reached may differ from those who responded. These results are best read as estimates, not exact measurements.
Full methods report (PDF) Recruitment survey methods (PDF)