RESEARCH STUDY Digital Recovery NEAR Labs at University of Pittsburgh
Digital spaces can help or hinder addiction recovery. At the NEAR Lab (University of Pittsburgh), we study how digital life overlaps with recovery: how people give and receive support in forums, chatrooms, and virtual meetings; how AI changes those conversations; and how shared culture (even popular music) shapes recovery journeys.
Our research points to something powerful and simple: supporting others keeps people engaged. From my earlier PhD work: new online community members who replied to others’ posts early on, had about 30× higher odds of staying active over three months. That’s a significant number. Other studies support this idea: people tend to do better when they’re active in recovery communities—online and offline.
We’re digging into which kinds of support-giving matter and how. Emotional support (encouragement, empathy) is quick and accessible. Appraisal support (constructive feedback) takes more care and focuses on someone’s behavior or situation. Informational support is important too. We’re studying how humans can get better at being supportive. Meanwhile, we’re closely tracking AI’s growing impact on these dynamics.
While a growing number of research studies focus on how people can extract “recovery capital” from communities, our focus is on what people can contribute—because giving support is changeable, learnable, and scalable. We want to understand how giving support strengthens individual recovery and the health of recovery communities.
If you’ve used online communities to change your relationship with alcohol or other drugs, your perspective matters. We’re building a waitlist for upcoming studies (brief surveys and occasional focus groups). Tap below to learn more and sign up for updates.
Learn more about current studies here:
https://open.substack.com/pub/onlinerecovery/p/recruiting-for-online-survey-study
Thanks for your support!
Dr. Jason B. Colditz
Networks Enhancing Addiction Recovery (NEAR) Lab | University of Pittsburgh

