‘I didn’t flourish’: Evergreen study participants reflect on using the wellness app during its first campus-wide test
Four Evergreen study participants were granted anonymity to speak about their experiences due to concerns about study eligibility. They have been given the pseudonyms Evan, Gus, James and Sarah. Evergreen, the campus wellness artificial intelligence project developed by Dartmouth professors and students, recruited 200 undergraduate students for a research study conducted over the course of the fall 2026 term. The Dartmouth followed up with participants as the study nears its conclusion. At the “AI at 70” conference on Oct. 30, Evergreen founding director Lisa Marsch announced that the initiative will conduct an expanded study next year to recruit 600 undergraduates to “test the fully generative AI version that integrates sensing, full personalization and adapting interventions.” The version of the app used in the fall study does not use generative AI, and instead features a “chatbot” with “pre-written” messages, Marsch previously told The Dartmouth. The study compares a control group and an experimental group by their “flourishing,” defined by Evergreen as the “holistic state of well-being where a student thrives across multiple dimensions of life,” according to the study’s consent form obtained and reviewed by The Dartmouth. The control group received a self-paced “flourishing guide” intended to be read and used on the participant’s own schedule. The experimental group received a smartwatch — an Apple Watch SE 3 or a Garmin vívosmart 5 — provided by Evergreen for the duration of the study and downloaded the Evergreen app on their phones. The app and watch tracked and provided the user’s biometric and technological data, including heart rate, light exposure, physical activity, sleep patterns, location, frequency of phone calls and text messages. The app was also synced with each participant’s calendar, Dartmouth Dining, Canvas and DartHub data. Participants opted into providing each source of data. Each participant in the fall study will be paid $300 over the course of the term, totalling approximately $60,000 across for participant compensation. Students are paid in three installments, following their onboarding, a midpoint survey and a final interview. Evergreen is privately funded through “philanthropy and largely alumni and parents,” Marsch wrote in a Nov. 16 email statement to The Dartmouth. Sarah said she felt like she was being paid “way too much” for her effort because she used the app “maybe once a week” and would have participated in the study for “only $25.” Sarah added that she knows other participants who are using the app more frequently than she is but are “just clicking through the prompts and not reading them.” “They’re just going through it as fast as they can, because you don’t have to really pay attention,” she said. “You don’t have to read any of this stuff for it to say that you’ve completed this [the modules].” Sarah added that she did “not really” find the app helpful but followed “one” recommendation from the chatbot to wake up at the same time every day. Gus said he similarly used the app for “maybe five minutes a week” each week. “I would get the notifications all the time, and I would look at the notifications and I would think, ‘That does sound slightly helpful,’” he said. “There was nothing holding me to it, though, so I just wouldn’t do it.” To improve the application, Gus said that he believes there should be some “sort of incentive to help me find out why I want to use this app.” He contrasted his experience on the Evergreen app with the human interactions he had when he needed advice about his grade in a class. “I emailed the undergraduate dean and the Academic Skills Center because they actually had people that could hold me accountable,” he said. “Just by the nature of scheduling a meeting with them, I had an incentive to show up at the very least and improve myself, because now, there’s expectations on the line.” Sarah said she felt how the chatbot’s prewritten chats were not “interactive or tailored to me at all.” Sarah also said the onboarding process made her believe the app was “more developed than it is.” “I think they’re trying their best to make this a good research study,” she said. “But the app just isn’t engaging. They aren’t at a point yet where they can really get the most out of research participants.” James called the information provided by prewritten chats in the app as “common sense.” “It just feels unnecessary,” he said. “I can definitely imagine it being helpful to some people — it can be a rock. I just feel like I personally don’t really need that.” Evan, who is in the experimental group, said the app “has not really changed” his “Dartmouth experience.” He thought the Evergreen team would give “more guidance” on how to use the app. Even after completing the trial, he still does not feel like he knows how to use the insights page, which tracks and reports sleep and physical activity. His page showed zero movement and sleep in the last month, according to a screenshot obtained and reviewed by The Dartmouth. Evan also said he “never” used the “mood tracker” because it was never explained to him during the in-app tutorial. Evan said it was “interesting” to see that the Evergreen team was prioritizing questions about whether using the app was “addicting” and whether it increased his phone usage in the end of term survey. “It’s not an addictive thing, but I guess they’re worried about being seen that way,” he said. Sarah said the “only drawback” about participating was “thinking about how much money” went into the study. “For me, it feels like a very minor project that they’re investing a lot of money into,” she explained. Gus said the app’s notifications and reminders were “helpful” because they gave him a “passive reminder” that he should be “doing something to improve [his] life.” The experience “wasn’t overwhelmingly positive; it was slightly helpful, and that was about it,” he said. When thinking back to the start of the study, Gus reflected how Evergreen described itself as an app to “help students flourish.” “I thought I would flourish, and then, I don’t know, I didn’t flourish,” he concluded. Gus said he believes that the app will be “much better” once it incorporates generative AI. The prewritten scripts feel “too limited,” he explained. James said he does not see himself participating in the 2027 Evergreen study that will incorporate generative AI because he believes generative AI “isolates people.” “It has done bad things before like drawing people further into psychosis, and it’s just not something I’m comfortable with,” he said. “Even if I don’t think I’m susceptible to it, I don’t want to legitimize it by participating in the study.” Clara Thorsen Clara Thorsen ’30 is a reporter from Seattle. Outside The D, she enjoys hiking and filmmaking. She can be reached at [email protected]. Isabela Pierry Isabela Pierry ’29 is a reporter from New York and is majoring in comparative literature and government.