Probably too late to change it, but for the question "how did you hear about this survey?", the EA Forum isn't mentioned :P

Thanks for doing this!

Under "Have you received career coaching from 80,000 Hours?" there are 3 options: "I have received career coaching", "I have not received career coaching, but would like to", and "None of the Above". I think if "None of the Above" was replaced by "I have not received career coaching, and would not like to" then you'd more accurately measure people in that category.

IMO the EA survey is a super powerful tool that's currently underused. Here's an idea bank for future surveys:

• Ask what skillsets people in the community are attempting to build, and what career paths they are trying to move into. Maybe we can forecast talent gaps in advance and build/recruit for those skills, or identify if there's a glut of people moving into a particular area. Bonus: In order to help people coordinate to avoid gluts, also ask people how dedicated they are to their current career path/how much career capital they've built. Then if I'm in an overpopulated career area, and I know I have less career capital for this area than the average, I know I'm one of the people who is best-positioned to move out of it. (You might even set people from the survey up with each other in order to overcome coordination challenges of this type.)

• A lightweight method for facilitating comparative advantage trades: In addition to asking people what career they are personally working on, also ask them what careers they think more EAs should work on. Then have EAs who are just getting started with the movement and feeling directionless look over the freeform responses for ideas. That way I can continue in a career path I have comparative advantage for while still getting to influence how our collective career capital is allocated on the margin.

• You could also ask people if they are open to being contacted by EA organizations that are recruiting for their skills. 80K says talent gaps are big and junior hires are valued at over $1M by EA orgs. I'm guessing a lot of hires currently happen through networking, which is a relatively inefficient process. Using the EA survey as a talent clearinghouse could generate millions of dollars of value on an annual basis. I assume you'd first want to talk to EA orgs to see if a process like this might work for them. I can think of a few advantages of this relative to using LinkedIn: career profiles optimized for what EA orgs are interested in, avoid sketchiness of unsolicited LinkedIn messages, probably a more comprehensive and up-to-date database of potential hires. Also, I remember CEA complaining that a lot of people are interested in EA, but it seems harder to get people seriously involved. Maybe sending emails of the form "EA organization X wants to hire you!" could help overcome inertia. You could still use mutual connections on FB/LinkedIn to measure involvement & dig up references. One complication is you'd want to separate the survey into "professional" and "personal" sections to control what information potential employers see, but I think the potential upside is worth it.

• Add calibration questions.

• Ask people which causes they've changed their minds about and why.

• Ask EAs about their biggest productivity bottlenecks.

• Ask people what mental health issues they suffer from.

• Ask people how much $ they have in donor-advised funds etc. that they are saving up for future giving opportunities, and what circumstances would trigger donation. In general, it'd be nice to know how the community as a whole currently balances giving now vs giving later. Asking people about the circumstances that would cause them to donate could also help unendorsed donation procrastination.

• How many people read/contribute to online EA discussions? Why or why not?

• What factors are holding people back from being more involved in EA? Why do people choose not to work for EA organizations?

• LW and SSC surveys might have more ideas. (A number of the above ideas are things I remember from the LW survey that I wish the EA survey had.)

I was planning to give some feedback on the 2017 survey instrument after the last post in that series, which I had assumed would finish before the 2018 survey was released. Since my assumption was wrong (sorry!), I'll just post my feedback here to be considered for the 2019 survey:

1. One major aspect of EA is the regularly produced online content on this forum and elsewhere. It might be useful to ask about the average number of hours a week people spend reading EA content as that could help people evaluate the value of producing online content.

2. You could also ask people whether they've attended an EA Global conference. The responses could be used as a proxy to distinguish more involved and less involved EAs, which could be used in analyzing other issues like cause area preferences.

3. For the question about career path, you could add advocacy as a fourth option. (80,000 Hours treats it as one of the four broad options.)

4. For the same reasons that race was included in the 2017 survey, it could be useful to ask about parental education (as a proxy for socioeconomic background).

5. You could ask people how many of their acquaintances they have seriously attempted to persuade to join EA and how many of those did join. This could provide useful data on the effectiveness of personal outreach.

6. Another question that may be worth asking: "Have you ever seriously considered leaving EA?" For those that answer yes, you could ask them for their reasons.

7. I think it could be useful to have data on the percent of EAs who are living organ donors and the percent of EAs who intend to become living organ donors. The major downside is that it may cause people to think that being a living organ donor is part of EA.

8. Borrowing from Peter Singer, I propose asking: "Has effective altruism given you a greater sense of meaning and purpose in your life?"

9. You could also ask about systemic change: "How much do you think the EA community currently focuses on systemic change (on a scale of 1 to 10)?" and "How much do you think the EA community should focus on systemic change (on a scale of 1 to 10)?" You could include a box for people to explain their answers.

10. Lastly, you could ask questions about values. A) "Do you believe that preventing the suffering of a person living in your own country is more important than preventing an equal amount of suffering of a person living in a different country? Assume that there is no instrumental value to preventing the suffering of either and that in both cases the suffering is being prevented by means other than preventing existence or causing death." B) "Do you believe that preventing the suffering of a human is more important than preventing an equal amount of suffering of a non-human animal? Assume that there is no instrumental value to preventing the suffering of either and that in both cases the suffering is being prevented by means other than preventing existence or causing death." C) "Do you believe that preventing the suffering of a person living in the present is more important than preventing an equal amount of suffering of a person living several centuries from now? Assume that there is no instrumental value to preventing the suffering of either and that in both cases the suffering is being prevented by means other than preventing existence or causing death." D) "Do you believe that it is bad if a person who would live a happy life is not brought into existence?"

The EA community climate survey linked in the EA survey has some methodological problems. When academics study sexual harassment and assault, it's generally agreed upon that one should describe specific acts (e.g. "has anyone ever made you have vaginal, oral, or anal sex against your will using force or a threat of force?") rather than vague terms like harassment or assault. People typically disagree on what harassment and assault mean, and many people choose not to conceptualize their experiences as harassment or assault. (This is particularly true for men, since many people believe that men by definition can't be victims of sexual harassment or assault.) Similarly, few people will admit to perpetrating harassment or assault, but more people will admit to (for example) touching someone on the breasts, buttocks, or genitals against their will.

I'd also suggest using a content warning before asking people about potentially traumatic experiences.

For the (bonus) question on comparing impact of an "average" charity with the best one, it would probably be worth specifying whether we mean mean or median, as I know some people here expect (I think myself included) a very significant difference between mean and median.