Thanks for writing this! I’m curious about how important you think the selection effects are here? eg. in terms of people who have had worse experiences with EA being less likely to fill out the survey.
I think my initial guess here would be that EA is roughly net neutral for people in terms of mental health, and because I expect the selection effects in who takes the survey to be fairly strong, I don’t really update much on seeing the above data
>my initial guess here would be that EA is roughly net neutral Yes, I think the null hypothesis is a good starting point. Sometimes I see people starting from the guess that EA has a bad effect, and that’s not what the evidence suggests.
I do think it is likely that many people get exposed to EA and realise it isn’t a great space for them personally, and choose not to engage as a result. These people won’t have filled in the EA Sometimes those people might not be a good fit for EA anyway—they have different priorities or values in their lives. But I expect some would have been very value aligned and gotten a lot out of being involved, so that is the group I’m worried about.
The Rethink Priorities data makes me (tentatively) a little less worried that a large number of people spend a couple of years engaging EA (and so fill in the survey) and find it too rough for them and then leave (and don’t fill in subsequent surveys). I DO think this happens to some people. But if this was a very large pattern I would expect community member who had been around longer to say it is more positive for their mental health, whereas the data points gently in the opposite direction.
I don’t know whether active community members are more or less likely to fill in the survey if they’re having a rough time in EA. I think it could go either way.
Thanks for writing this! I’m curious about how important you think the selection effects are here? eg. in terms of people who have had worse experiences with EA being less likely to fill out the survey.
I think my initial guess here would be that EA is roughly net neutral for people in terms of mental health, and because I expect the selection effects in who takes the survey to be fairly strong, I don’t really update much on seeing the above data
Seems plausible to me that people who are interested in the topic of mental health have worse mental health than the average person.
>my initial guess here would be that EA is roughly net neutral
Yes, I think the null hypothesis is a good starting point. Sometimes I see people starting from the guess that EA has a bad effect, and that’s not what the evidence suggests.
I’m really not sure.
I do think it is likely that many people get exposed to EA and realise it isn’t a great space for them personally, and choose not to engage as a result. These people won’t have filled in the EA Sometimes those people might not be a good fit for EA anyway—they have different priorities or values in their lives. But I expect some would have been very value aligned and gotten a lot out of being involved, so that is the group I’m worried about.
The Rethink Priorities data makes me (tentatively) a little less worried that a large number of people spend a couple of years engaging EA (and so fill in the survey) and find it too rough for them and then leave (and don’t fill in subsequent surveys). I DO think this happens to some people. But if this was a very large pattern I would expect community member who had been around longer to say it is more positive for their mental health, whereas the data points gently in the opposite direction.
I don’t know whether active community members are more or less likely to fill in the survey if they’re having a rough time in EA. I think it could go either way.