Introduction
Overall, I had a very positive experience in CS247B this quarter. I really enjoyed working with such a great team, and was genuinely excited to explore our behavior change (bad posture). Before this class, I thought of behavior change through a much more simple lens. I saw it as just about providing people with the right information and incentives. Now, after learning about nudges, anchoring, and conducting our own intervention and user studies throughout the course, I realize there is a lot more to consider about behavior change. Changing someone’s behavior is not as simple as I initially thought.
Work Experiences
Throughout the project, I was able to incorporate my previous experiences in human-computer interaction. I had worked with digital applications before, so I was familiar with designing interfaces, but this class introduced me to deeper considerations around behavior change techniques.Working on Postura, our posture correction app, challenged me to think critically about motivation, social influence, and ethical nudging in design.
I loved working with my group! All of the activities during class, from the ethical discussions to brainstorming about our project, were very enjoyable and fulfilling. I think we had a good balance of sharing all our opinions and also being able to come to an agreement on one strategy. All my team members were also very supportive and created a fun team environment, which made working on the project much more enjoyable. My favorite assignment was the moodboard, because it gave us the opportunity to think about the personality of our solution and get inspiration. It was also fun to create something tangible together, and I think we had a very cool result.
Our moodboard!
We encountered a few issues related to general user confusion in our assumption tests and usability tests. Most of our issues surrounded our flow for rewards. With Postura, there are several different possibilities for rewards, including an individual reward, group reward, or no reward at all. There is also a prize pool every reward period that all users contribute to, and only the users who reach the specified goal get to split the prize money. In addition, our solution requires users to purchase a companion physical device in order to track their posture. The idea of spending money, especially when they did not expect to, seemed to scare many users. In our final prototype, we add an extensive tutorial flow for first time users that lays out how exactly Postura works and gives a rationale for why it works this way. This also ensures that we are completely transparent and up front with users, and that they know exactly what they are signing up for.
As a redesign for this project, in order for users to stay engaged over a longer period of time, I might try to explore incorporating more intrinsic motivation, since most of the strategies currently engaging users are extrinsic (gamification, rewards). It might be interesting to explore progress tracking over a longer period of time more deeply to see what methods resonate with users.
Ethical Considerations
We iterated extensively on our vision for Postura. Initially, our concept still involved creating a group to track your posture with friends and a leaderboard to see everyone’s progress. However, instead of an automatic tracker, friends would take photos of each other and post it to the group. For every photo with bad posture, the person would lose a certain amount of points. Whoever came in last place every week would have an automatically generated photo with a funny filter on it. While this idea initially seemed lighthearted, fun, and likely to encourage friendly competition, we also saw many issues with this. There are privacy issues associated with people’s photos being posted on the app. We ran into questions like: What if the photo is inappropriate? What if the user did not give their consent for that specific photo to be taken? How do we balance fairness in terms of the point system while also honoring the user’s wishes for which photos of theirs are posted? This method also encouraged more negative reinforcement than positive, and could lead the user who came in last to feel ostracized and bullied.
Our current solution relies on positive reinforcement. Since in our current system, users can receive rewards, we plan to use friendly reminders rather than guilt tripping. Some nudges/cues we currently use is an exclamation point next to the groups icon when there is a reward to be claimed. If a user skips and decides not to join the current round, they can still see how their friends are doing, including the leaderboard and general group progress. This is a subtle nudge, since we were hoping that seeing their friends participate would make users more likely to change their mind and join the group. We justified these decisions since they lead to users having better posture, which is the goal of our intervention.
Privacy is very important, since we are storing users’ posture correction history. We plan to store all user data very securely, using encrypted identifiers that would not easily identify the user it belongs to. Users also have complete control over their account. If they delete their account, we delete all data associated with their account.
Next time, when faced with a similar situation I will make sure to establish team norms at the very beginning, because I think that really helped with team communication. I think it might help to think about the ethical considerations and overall vision for the app before discussing what specific features we might implement and the actual functionality. That may have helped reduce the number of iterations we had when finalizing our idea. In addition, during the usability testing I learned how important it is to evaluate prototypes from the perspective of a first time user, because there were many flows that were confusing to other students but not to us since we had been working with the prototype for a couple of weeks.
