Yoonsang’s poster linked influencer tweets to real-world cigar sales. An increase in tweets from high-follower and high-mention accounts was associated with a significant rise in weekly unit sales (~55k to ~220k more units, respectively). #PublicHealth #HealthComm

Miao's poster looked at anti-vaping campaign opposition. On Twitter (X), it’s often organized & political. On TikTok, it’s more decentralized, humorous, and ironic, showing how platform culture shapes public health messaging.

Props to our own Alex Kresovich on this study, with JCOIN project leadership from Bruce Taylor and Phoebe Lamuda!

Read the full paper: www.tandfonline.com/doi/full/10....

The takeaway: To be effective, messaging must meet audiences where they are.

The biggest twist? When people with higher stigma had used opioids themselves without developing an addiction, it reinforced their belief that OUD is a personal failure, rather than building empathy.

The study found that people with high vs. low stigma don't just disagree—they operate in completely different realities.

➡️ High-stigma view: OUD is a personal choice. ➡️ Low-stigma view: OUD is a medical condition shaped by societal causes.

Final insight for community managers: A small group of "super-users" (just 4.55% of chatters) can have a massive impact, driving nearly a quarter of all chat messages!

Kudos to the SDC's Alex Kresovich & Simon Page for leading this analysis! #DataScience

Finding #2: Content format is everything. 📺

In-Real-Life (IRL) streams sparked the most STEM talk (21.65%) - far more than gaming streams. Also, streams with lots of peer support were over 3 times more likely to have a super-active chat.

Finding #1: Your community leads the learning. 🧑‍🏫

Regular viewers drove MORE STEM discussion than the official team members (16.9% vs 11.3%). True engagement isn't top-down; it's peer-to-peer.

#OnlineLearning

How did we find this? By pioneering an ethical data donation method to reveal what people see online. This shows how ad targeting may exacerbate difference in health outcomes. Lead author: Yoonsang Kim. #HealthEquity #DataScience (4/4)

But on Instagram, the story changes. There, Black users faced the highest food & beverage ad exposure (25%), while Hispanic users saw less exposure (10%) than White users (16%). (3/4)

On Facebook, the differences were sharp. Hispanic users were exposed to food & beverage ads at more than double the rate of White users (33% vs 15%). The rate for Black users was 22%. (2/4)

🎶 What's your go-to productivity hack?

“Music and coffee! I also try to stop working when things are still interesting, so I’m naturally pulled back in the next day to keep the momentum going.”

🍜 What's a fun fact about you?

“I was born in Sichuan, a region in China known for its bold, spicy flavors. If I ever switched careers, I’d probably open a noodle shop. Making noodles with chili oil and Sichuan peppercorn is something I genuinely love.”

💡 What's a current challenge you're excited to tackle?

"Through my NCI-funded project, I’m analyzing a decade of e-cigarette prevention campaigns across X and TikTok. The challenge is making sense of the patterns to better understand what campaign resilience looks like in today’s environment."

💬 What’s your favorite part of working in the SDC?

“I’ve been with the SDC since 2014, and what continues to inspire me is how we approach research: with rigor, creativity, and a willingness to evolve. We work across disciplines and methods, which keeps the work dynamic and deeply rewarding.”