Here’s as much insight into their method as ChangeResearch release:

Change Research conducted a poll of 1,203 2020 Trump voters nationally on November 18-21, 2020. The full 2,559 voter sample was weighted to reflect the 2020 electorate, and the results were filtered to those who reported voting for Trump in 2020. All data shown in this document is among Trump voters only and the margin of error is +/- 2.83. Change Research reaches voters via targeted online ads that point people to an online survey instrument. Our Dynamic Online Sampling establishes and continuously rebalances advertising targets across region, age, gender, race, and partisanship to dynamically deliver large samples that accurately reflect the demographics of a population. Post stratification was done on state, gender, age, race, education, and 2020 presidential vote.

Selecting online only people, and through ads no less, is going to bias the replies to the online-addicted.

I’m betting many people who voted for Trump never will see those request for participation from ChangeResearch.