June 16, 2023
Poll contact: Tatishe Nteta
Majority of Americans Support Supreme Court Reforms, Including Term Limits and Ethics Requirements, According to New UMass Amherst Poll
Poll finds overall disappointment in the performance of the court and Congress, division over the future of affirmative action
YouGov interviewed 1,298 respondents, including 1,165 main sample respondents, and an oversample of 133 African Americans. The main sample was matched down to a set of 1,000, and then combined with the oversample to form a final dataset of 1,133 respondents. The main sample was matched to a sampling frames on gender, age, race, and education. The sampling frame is a politically representative "modeled frame" of US adults, based upon the American Community Survey (ACS) public use microdata file, public voter file records, the 2020 Current Population Survey (CPS) Voting and Registration supplements, the 2020 National Election Pool (NEP) exit poll, and the 2020 CES surveys, including demographics and 2020 presidential vote.
The matched main sample and the unmatched oversample were then weighted to their respective frame using propensity scores. The frame used for the oversample is similar to the previously described frame from the main sample, with the main difference being that it only African Americans. The matched cases and the frame were combined, and a logistic regression was estimated for inclusion in the frame. The propensity score functions both included age, gender, race/ethnicity, education, and region. The propensity scores were grouped into deciles of the estimated propensity score in the frame, and post-stratified according to these deciles.The weights for the main sample were then post-stratified on 2020 Presidential vote choice, followed by a four-way stratification of gender, age (4-categories),race (4-categories), and education (4-categories). Meanwhile, the weights for the oversample were post-stratified on 2020 Presidential vote choice, a three-way stratification between gender, age (4-categories), and education (4-categories), and finally an individual stratification on region.
Then, the matched and weighted main sample and oversample datasets were combined into one. From there, the proportion of African Americans was weighted down to produce the final combined weight. After that, a subset of this combined dataset was taken so that only observations involving African Americans remained. This AA subset of the combined dataset was weighted to the same frame as the oversample using propensity scores. The unmatched cases and the frame were combined, and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity, education, and region. The propensity scores were grouped into deciles of the estimated propensity score in the frame, and post-stratified according to these deciles.
Finally, the weights were then post-stratified on 2020 Presidential vote choice, followed by the same three-way stratification mentioned earlier, in order to produce the final AA weight.
Field Dates: May 31 – June 8, 2023 | Sample: 1,133 Respondents | Margin of Error: 3.4%