[elpool2]

Is it an ethical breach if I don't get informed consent from users when I do A/B testing on my website? Or is it only unethical if I publish the results afterwards? Or is it only if I'm actually trying to affect the user's mood specifically?

[gemma]

If your A/B testing is intentionally inducing a negative mental state in your users without letting them know up front, I'd say that's probably an ethical breach.

In general, A/B testing is studying the website. You're not testing us, you're testing the best ways to convince us to do stuff. On top of that, when users visit your website (which is presumably pitching something to them), they know they're being pitched. They know the website is going to try to convince them to do stuff--click over here, watch this video, sign up for foogadgetmcawesome. Same drill with advertising. Yeah, this commercial with the freaking Budweiser puppy made me cr--er, chop onions--in its attempt to get me to buy beer, but I know they're trying to sell me beer, and I knew as soon as the puppy hit the screen that I'd probably start sniffling.

[elpool2]

That's a good point about the distinction between studying the website and studying people, but I'm not sure that's really the line that was crossed. If the study had asked "Does this newsfeed algorithm make people sad?" instead of "Do positive posts from friends make people sad?", would that then be ethical? It's still manipulating people the same way, and I think people would still find it creepy. What if it was the exact same algorithm, but instead they asked "Does this make people click more ads?", is it now an ethical A/B test just because the intent is different?

[zaroth]

A/B tests look for user response to stimulus or change. This study certainly did nothing more. The ethics of a deliberate action can certainly turn on its purpose, but I strongly doubt this is such a case.

[benastan]

I don't see the distinction. It's fine to market your product, but there is such a thing as distortion and as a scam. Providing a fair representation of your product factors into ethics, as does Facebook fairly presenting the feed of your friends news, with a reasonable attempt at NOT distorting it.

[gemma]

I agree! I'd put unethical behavior like distortion and lying in a different category than Facebook's, but I agree. I was trying to kill the parallel between Facebook intentionally manipulating user emotions for science, and Budweiser intentionally manipulating viewer emotions for beer sales. Video games work here too: Amnesia scares the unmentionables out of me, but I'm ok with that, because I know that's the whole point.

Another missing component here is user control. I can turn off the commercial and stop playing the game. I'm currently working (in a minor capacity) on the National Children's Study, and our participants can walk away at any time. When the Facebook subjects on the negative side started feeling slightly more sad, they had no idea why, and no clue how to stop it.

[cududa]

This is where the real discussion needs to focus. These are the hard questions we need to consider

[gammarator]

A/B tests are different, though: in aggregate, they should make your website better for me, the user. Here the user is just fodder for an academic's fairly trivial journal article (my friends' feelings affect my feelings!).

Who decides what experiments social scientists get to run on 1/7th of the world's population?

[akerl_]

They only make the site better for you if you're on the "winning" side. If A/B tests made the site better for everyone being tested, they'd not be very useful. Unless you mean "they should make your website better for me [after the test is done]", in which case they have the same goal as Facebook's study.

[gammarator]

That's what I said: A/B tests improve the site "in aggregate."

The goal of this study was not to improve the site, though: it was to test a hypothesis about social psychology.

[yincrash]

The goal of the study _is_ to improve the site. Facebook wants the users to return to the site, and users who become happier when using the site do that. Facebook isn't paying its data scientists to perform studies that have no meaningful use.

[akerl_]

I sincerely hope you're not making the case that testing a hypothesis for a scientific paper is not a valid reason to perform such an experiment, but improving a website is.

[jacquesm]

That depends on the goal of your A/B test. I can see a ton of possibilities where A/B testing could be used for un-ethical purposes.