Let's talk a little bit more about the ethics of A/B testing. It's not a subject that a lot of people think about very often, but it matters because the people who are using your sites are being experimented upon, and they don't even know it. So, we're all unknowing guinea pigs. On the websites that we use commonly every day, there are often A/B tests that are going on. In the user's mind, they aren't sure what's real because what they're experiencing is what's real to them. Even though you might be testing an A version, a B version, a C version, their lived experience is just that version and so that is the experience that is most important to them. You have to give some thought to what's appropriate to test with them, and also about whether or not it's something that you would feel comfortable experiencing if you were on the other side of this equation. All digital products have some sort of security privacy in usage agreements that gives some broad outlines around what is permissible usage of the site, and set expectations on users part about the information that's being collected about them, and as well the things that they are expecting from you in terms of protections and privacy. A/B testing is conducted within the confines of those agreements. So, you really have to think about whether or not the things that you're trying to do are appropriate. Given the context of what you're working on that can take on some pretty significant implications. So, a few things to think about when you are structuring an A/B test. In some circumstances, it's actually really important that you take into consideration how people are going to think and feel based on what it is that you are presenting to them. Because they can't opt out. So, obviously the intent that you have going into creating a variant matters. Are you really trying to get people to do more or less of A or B? Do you feel comfortable getting people to do more or less of A or B? Finally, one of the things that's a little bit different about doing A/B testing on a live site is that you don't have an institutional review board typically that's looking over your shoulder to see whether or not what you're doing is okay. Like you would in a typical lab setting. So, a lot of those rules that would be imposed fly out the window. So, even if you aren't having that guidance from an Institutional Review Board, always let your conscience be your guide. Because it's really important to make sure that you feel comfortable about the choices that you're making for and on behalf of users, so that you are doing things that are going to help and not hurt them. A couple of years ago, Facebook revealed publicly that they had done some A/B testing with users with over a half million users in order to adjust the number of positive and negative posts that they saw in their feed. It was all part of a psychological experiment to see how emotion spreads via social media. So, Facebook wasn't trying to manipulate the feeds make people feel bad or better, they were just really trying to understand what is the impact of the mix of positive versus negative posts in a feed on how emotions spreads via the Social Web. If you would like to read this article in further detail, there is a link provided. So, there are a couple of ways that you can do research. There's the do no harm approach, where you're really trying to help people, and then there's the approach where we're really just trying to measure understand, quantify, dig deep. Part of the problem with this is that if you're doing an A/B test where you're just measuring to see what's happening, one of the unintended consequences of doing this type of research is that you might actually negatively impact people. One of the things that they will not ever know is the fact that they're being experimented on in the first place because it's part of an A/B test. So, if you're testing to understand measure something that can actually have an impact on people's feelings, and the way that they live, and they don't even know about it, it can make people feel really uncomfortable. So, when you're doing research A/B testing since people aren't opting into doing any of these tests, you really do have to think about practically what the implications are of, the things that you're doing in order to change their experience. Is it really going to make it better? All of the A/B testing out there for the most part is about optimizing hopefully positive impacts in order to make an experience better. There are actually on experiences that are not about making people feel better, do better, be better. When you're talking about ethics of UX, and you're talking about ethics of UX research, a lot of the time people are creating experiences now that promote addictive types of behaviors. So, you have to be very careful about how it is that you think about optimizing the experience in that situation. So, most of the time A/B tests are not visible to users. They don't really know that there is a version that they're seeing that looks different from a version that somebody who might be sitting right next to them at the very same time is seeing. Most of the time, that also is because we strive to make those differences be relatively minimal so that people aren't seeing that there are these huge differences between what it is that they are experiencing, and what others experience. However, sometimes those changes really can have a significant impact on how people think, feel or experience the world. You need to be thinking about the impact of those changes as you are planning and executing your A/B tests. You need to make sure that you are comfortable with the choices that you're making, and that if you put yourself in the shoes of the users that you are testing with, that you would feel comfortable with the impact of the things that you're exposing people to.
