Before this class I think I pretty much just thought that behavior change was hard. I don’t think I considered or realized all of the different ways that behavior can change. Changing behavior to align with what you want your behavior to be is a very small subsection of behavior change that is especially hard to harness, but behavior change itself happens all the time. One of the biggest things that our instructor Christina emphasized when finding the people for user studies was to find people who were motivated to change. I think it’s really easy to find people that want to change, but really hard to find people motivated to change. The same thing applies on the individual level—it’s really easy to find something that I want to change, but hard to find the motivation to actually make that change. With the rise of social media and doom scrolling, people (myself included) build habits they never intended because companies use extremely effective manipulation tactics, especially when it comes to hijacking your reward system. I’m specifically thinking about how dopamine is released right before an unpredictable reward—the more unpredictable it is, the higher levels of dopamine. So when I’m scrolling through TikTok, at every scroll I might find something truly delightful—a reward—or I might find something totally uninteresting or bland. This unpredictability makes that habit really addictive, and it’s not something that I chose. One thing I wished I did more in this class is try to harness this same power for good. But onto the stuff I did do…
The diary study was interesting but challenging. It was really hard balancing asking the people in the study enough questions to get quality info without scaring them off with too much of a burden. I would have loved to include a lot more questions and get a lot more details from them, but it’s ultimately up the the person in the study how much they’re going to write. When people wrote a lot on the open-ended questions, we were able to glean the most interesting and insightful information that had the biggest impact on our design, but the vast majority of people filled out the bare minimum. I can’t blame them because if I were in their shoes, I think I would too. I’m taking CS377U Understanding Users next quarter where I will continue to create and run diary studies which will give me the opportunity to dig deeper into these ideas. I hope to get better at designing studies that balance participant burden with the depth of insights needed.
I also noticed that with our specific behavior (staying connected with friends after college), the people we had in our study were individuals that we thought to reach out to. It’s clearly a biased sample, and the people who would have been the most useful to the study because they’re experiencing the issue of isolation the most cannot be reached as easily. How can you mitigate this bias? How can you design for people who are less likely to be studied? Or will they just be left behind?
One of the most surprising insights came during our intervention design phase. Initially, we wanted our app to connect with users’ messages so we could automatically track when and from whom messages were received. We thought that since we weren’t asking for the literal contents of the messages, users would go for it. However, as we reflected on the implications of this request (and discovered that Apple has no way of allowing such access), we got feedback from users that they actually liked the process of logging their interactions because it gave them an opportunity to reflect. This insight proved our assumption wrong, and I’m glad it was uncovered because I think we came away with a much better outcome. We opted for a manual logging system that gave users more control and actually enhanced their experience rather than diminishing it. The final product was something that isn’t just monitoring you but an active tool that you have to engage with.
One conversation that really fascinated me was about gamification, and I think our intervention toed the line for what counted as gamification. It’s fascinating how imagined goals or achievements like artificial digital badges can have such a strong effect on people. I’m very wary of my own behavior when it comes to using Duolingo, which I used to try to remember as much Spanish as possible before taking my placement test. Even though I had an external goal, day to day I was only motivated by keeping my streak going. But to keep the streak, I only had to do one lesson a day, so that’s what I did–the bare minimum to satisfy that imagined goal.
It’s interesting how our solution compares to something like Duolingo. In our app, you don’t get reward badges or satisfying sounds when you water a plant. The indication is less completionist—what you get is a good-looking plant, not progress tracking or badge collecting. I don’t personally think that’s a reward that will take the place of the reward of communicating with friends because it’s so passive, but I could be wrong. It’s possible that people will just want to see their plants happy. I believe our design avoids becoming manipulative because the user is free to log an interaction even if it didn’t happen. There’s no accountability, and the only one seeing it is the user themselves. If they end up only caring about the health of the plants and they just come back to “water” them, I think it’ll be disconnected enough from their friendships that it won’t negatively impact them. It’s not like someone is going to engage in bland, surface level, transactional check-ins with their friends just so that they can be honest when the press the sun button to say they texted that friend. That said, I could be wrong, and it’s possible that the success indicator of a healthy plant is enough to rework how people value their friendships. Obviously, to truly assess this risk, further user testing over a longer period of time would be a must.
Overall, I truly loved this class, the teaching team, my project partners, and above all, Nina. I’m so grateful to have had this experience and I wish everyone happy behavior change!
