If this forum/EA has ethnographic biases, here are some suggestions

post by Nathan Young (nathan) · 2019-05-06T11:20:51.105Z · score: -3 (15 votes) · EA · GW · 7 comments

I read a post on here asking for an ethnography [EA · GW] likewise 80k hours' most recent "Our mistakes" piece points out their under-prioritising diversity.

Let us assume that an ethnography (which I guess is like a census) shows there are cultural/ethnic biases in our community. What could we do?

Listen. We could listen to those from underrepresented groups to try and understand why they didn't want to be part of EA/ this forum

Build a new voting algorithm which weighs votes in relation to a more representative sample. In the same way that polls have to be weighted to make the people polled match society we could reweight the forum so that if there are say, fewer women of colour, their votes would still comprise the same proportion of society that they do in real life. This would mean that the frontpage (which I guess is the top ranked posts) would show a more representative sample of all posts.

Try and guess why those biases exist and test those guesses. Why do we think this ethnographic bias exists. How can we test to see if that's actually the case? To suggest that a certain group enjoys discussing issue more is a poor explanation. Why do they? There must be a reason for the bias and I want to know what it is, so that if (as I think is likely) it is a bad reason, we can fix it.

Here are 80k hours' other suggestions:

Make a greater effort to source candidates from underrepresented groups, and to use trial work to evaluate candidates, rather than interviews, which are more biased.
Ask for advice from experts and community members.
Add examples from underrepresented groups to our online advice.
Get feedback on and reflect on ways to make our team culture more welcoming, and give each other feedback on the effect of our actions in this area.
Put additional priority on writing about career areas which are over 45% female among our target age ranges, such as biomedical research, psychology research, nursing, allied health, executive search, marketing, non-profits, and policy careers.
During our next round of board reform, we’ve found a highly qualified woman who we have asked to join.
Do standardised performance reviews, make salaries transparent within the team, and set them using a formula to reduce bias and barriers.
Have “any time” work hours and make it easy to remote work.
Implement standard HR policies to protect against discrimination and harassment. We adopted CEA’s paid maternity/paternity leave policy, which is generous by US standards.

7 comments

Comments sorted by top scores.

comment by Holly_Elmore · 2019-05-06T18:53:25.983Z · score: 30 (12 votes) · EA · GW

As the author of the original piece on ethnography [EA · GW], I just want to clarify that what I want out of it is an understanding of the social function of our beliefs, which can help us weigh whether or not they are true. It wouldn't be a diagnostic of what's wrong with EA. Even if it did indicate we are very homogeneous (big surprise), that wouldn't necessarily mean we should change.

Ethnographic study could well reveal that many of the beliefs that bind EAs together are alienating to non-whites or lots of women. We already know it's not super-welcoming/appealing to older people. But that doesn't necessarily mean EA should change. We may see easy ways to change to increase diversity and we may see that the things we'd have to change to get better diversity numbers are really core to the movement. Also, if we did try to change, I'd keep an open mind about the ways in which we change. These 80k recommendations are ways to make the workplace more equal, and generally a good idea, but they can't really be applied to the community itself.

An ethnography would just be more information about EA-- any changes we make would be motivated by our beliefs about what EA should be. I don't ever see EA being all things to all people, and it shouldn't try. It should try to do the most good. Increasing diversity could be part of doing the most good (imo for epistemic reasons), but I think we need to be careful not to get swept up in the diversity moment of popular culture or think that imperfect representation is an automatic condemnation.

comment by Nathan Young (nathan) · 2019-05-07T07:46:43.378Z · score: 2 (2 votes) · EA · GW

First of all, youch, people did not like this post. That's okay.

I think we need to be careful not to get swept up in the diversity moment of popular culture or think that imperfect representation is an automatic condemnation.

I think we agree. I would prefer to understand problems rather than kneejerk solve them. Perhaps I was unclear.

However I do think that changing voting to match profiling is an interesting option to see whether it would change the kind of content the bubbled up.

comment by Holly_Elmore · 2019-05-07T19:48:29.065Z · score: 8 (3 votes) · EA · GW
First of all, youch, people did not like this post. That's okay.

Aww, I'm sorry-- I didn't mean to sound harsh. I get very sensitive on this forum so I hate that I made you feel that way. I guess I was just really eager to clarify that diversity was not why I wanted an ethnography done and not considerate enough of the position you laid out.

I have a strong reaction against weighted voting on the basis of demographics, but it would definitely be interesting to see how it changed things.

comment by Nathan Young (nathan) · 2019-05-08T07:07:37.288Z · score: 8 (3 votes) · EA · GW
I hate that I made you feel that way.

No need to apologise, I didn't mean my original comment individually, more as a kind of "gee whiz" to how much the blog bombed in general. But as I say, that's okay, noone was unkind, they just didn't like what I wrote. I think it can be easy for communities like this to be very dog-eat-dog so I think a little vulnerability/honesty might go a long way. Recently I have learned when I am insecure enough to be tempted to "man up" it's often better to show vulnerability.

What is the issues of things voted against demographics, if I might understand. Let's say tall EAs want a slightly different thing than short EAs. Scaling comments by height as if it were a survey means that if we have a fewer than representative number of tall EAs their votes would get weighed as more. That would mean the top posts/comments would be more likely to contain things which appealed to them, since (if they comprised half the representative population) they would control half the weighted votes. So new tall EAs would visit a site closer in tone/culture to what they would enjoy.

I don't see why this would result in a less rational site, but if certain issues it turned out were culturally more important to short EAs, it would be good to notice that, rather than thinking it was about rationality.

Frankly, most counter positions seem to lead to "representative voting control by minority groups would lead to a worse site" and I don't understand why that is the case. If they lead to increased growth in EA in minority groups that seems a good thing.

So my slightly clunky analogy aside, what do you think?

comment by Holly_Elmore · 2019-05-06T18:26:09.044Z · score: 20 (9 votes) · EA · GW
an ethnography (which I guess is like a census)

It's not, and it's not like a survey, either. It's when a skilled anthropologist, who understands many general principles of human culture, becomes immersed in a culture for the purposes of detailed observation and writes down their findings. The written account is an ethnogrpahy and the entire practice comprises the field of ethnography. The wikipedia article is accurate.

comment by vaidehi_agarwalla · 2019-05-06T18:14:26.911Z · score: 16 (7 votes) · EA · GW

Just a question about your phrasing - did you mean to say cultural/ethnic biases instead of ethnographic biases? Because an ethnographic bias would refer to the ethnographer's own biases when conducting the study, and I don't think that's what you're referring to.

comment by Nathan Young (nathan) · 2019-05-07T07:40:55.078Z · score: 1 (1 votes) · EA · GW

Yes, you are correct.