I judge this to be a problematic criterion. See this comment, esp. starting with "To put this another way ...", for why I think so.

That comment makes a lot of sense. It depends what we use the criterion for. In the survey, it's to gather information, and it's for precisely this reason that I chose not to ask if people were 'EAs' in your loose sense - almost everyone would say yes. I'm curious as to what uses do you think the criterion's problematic for.

My contention is that there's a distinct separation between, on the one hand, the general idea that we should be altruistic (in whatever sense we decide is meaningful and useful) and that we should seek to optimize the effectiveness of our altruism, and on the other hand, the loose community of people who share certain values, certain approaches to ethics, etc. (as I outline in the above-linked comment), which are not necessarily causally or conceptually entangled with the former (more general) idea.

It's a matter of a degree, but in the EA context (which sets a high bar), I personally call people 'altruistic' if (but not only if) they've donated >=10% of a real income for over a year or they've consistently spent over an hour a week doing something they'd otherwise rather not do to help others.

My contention is that there's a distinct separation between, on the one hand, the general idea that we should be altruistic (in whatever sense we decide is meaningful and useful) and that we should seek to optimize the effectiveness of our altruism, and on the other hand, the loose community of people who share certain values, certain approaches to ethics, etc. (as I outline in the above-linked comment), which are not necessarily causally or conceptually entangled with the former (more general) idea.

That's right, if by 'conceptually entangled' you mean 'necessarily connected', or even 'commonly accepted by both groups of people'. For example, I believe utilitarianism's widely accepted by EAs (though the survey may show otherwise!), but not entangled with merely valuing altruism and the effectiveness of altruism.

This is problematic for various reasons, I think. I won't clutter this thread by starting a debate on those reasons (unless asked), but I think it's at least important (and relevant to endeavors like this survey) to recognize this distinction.

I see no harm in thread-cluttering, at least here - go for it.