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".
I had no problem. Concluding based on evidence is a form of believing. You're just being a prissy atheist. :-P Actually, I'm being a pedantic research scientist, but yeah ... The concern with the word belief is fair. A few other secular folk disliked that wording too. Keep in mind that these items were taken from previously validated psychological scales, so if I messed with the wording then the results become questionable as they are not exactly comparable to past findings. Thus, I had to keep things untouched. Additionally, the use of words like belief can be seen as a pragmatic decision for both the purposes of recruiting and for choosing questions that are easily understandable to a diverse population of mixed educational and intelligence levels. I'm dealing with a wide array of participants here, a one size fits all approach was the goal but I was aware that some items would not capture the nuances and be ideal. Thanks for participating