Methodology

The survey is based on YouGov interview data from a politically balanced, representative sample of 1,000 U.S. adults 18 years and older, based on 2023 American Community Survey data. It was fielded from March 19–27, 2026. The margin of error is +/-3.5. The dataset is matched and weighted to be representative across age, gender, race/ethnicity, education, region, and home ownership status, with additional post-stratification on 2024 presidential vote choice, age, gender, race, and education. Analyses involved comparing across differences in caregiving experience, party affiliation, and ideology.

Respondents were asked to select their own political viewpoint from among six choices: very liberal, liberal, moderate, conservative, very conservative, not sure. In this report, we use the same labels to be clear and transparent, so “conservative” respondents represent one distinct group and “very conservative” respondents another. We did not combine both groups, so the “conservative” label does not represent all self-identified conservative respondents, only those that did not choose to describe themselves as “very conservative.”