FWIW, I also think it’s plausible responses will be biased towards low numbers because people want to be avoid looking like they’re bragging, don’t want to contribute to people’s stress, etc.
(But to be clear, I’m not saying I expect those different sources of bias to cancel out—it seems hard to say what the net bias would be—and so also endorse the idea of giving more weight to an anonymous survey.)
FWIW, I also think it’s plausible responses will be biased towards low numbers because people want to be avoid looking like they’re bragging, don’t want to contribute to people’s stress, etc.
(But to be clear, I’m not saying I expect those different sources of bias to cancel out—it seems hard to say what the net bias would be—and so also endorse the idea of giving more weight to an anonymous survey.)