Most social science research about beliefs excludes atheist participant samples . . .

Biggest problem is the word "believe". Some of us "conclude, based on evidence", as opposed to "believe".
Many people (only secular folk, though) had that same issue with the wording. Ultimately, the word belief was chosen for either two reasons: 1) because it was most applicable to be understand by the broadest range of participants in the study who come from various ideological perspectives with various verbal intelligences, and 2) because the word was used in the psychological instruments that were standardized and validated upon such a wording. Thanks for taking it!