Follow
"We oversampled individuals who identify as LGBTQ+, Jewish, Muslim, African American and Hispanic/Latinx. For the oversampled target groups, responses were collected until at least 60 Americans were represented from each of those groups"
I see absolutely nothing wrong with this methodology /s
@matrix Weeeeeell, if you then take into account the ratio between the sample size and the real representation in population that is OK but you can't quite do a good double blind test like that