THE QUIZ IN YOUR FEED: ASSESSING MENTAL HEALTH KNOWLEDGE WHERE PEOPLE ACTUALLY SCROLL
- Jul 15
- 4 min read
Updated: 2 days ago
Week 10 · Blog Post 17 · Web 2.0-Based Learning and Performance · Summer 2026
AN ASSESSMENT THAT DOESN'T FEEL LIKE ONE
I read a review by Alonzo and colleagues (2023) about using social media for assessment, and one line rearranged my week: most online assessment platforms are built for one-way interaction, but assessment on social media is primarily formative — dialogic, interactive, feedback flowing in both directions. Several of the studies they reviewed used a structure I can't stop thinking about: a pre-test, an intervention delivered through the platform, and a post-test. A whole miniature course, living where people already scroll.
My first reaction was skepticism. Would anyone actually spend that much time on a single social media post? I barely finish the carousels I saved on purpose. But the more I sat with it, the more I suspected the question isn't whether people will complete a test in their feed — it's whether we can design assessment so it doesn't feel like a test at all.
WHY MENTAL HEALTH IS THE RIGHT TEST CASE
So which topics deserve this treatment? I keep landing on mental health knowledge.
Mental health is a misconception-rich domain. Most people carry around confident, wrong beliefs — that asking someone about suicidal thoughts plants the idea, that depression is visible, that "just exercise" is a treatment plan. That's exactly the kind of knowledge a two-minute myth-or-fact quiz can surface and correct. It also chunks beautifully: warning signs, how to check in on a friend, what to say (and what not to say), where crisis resources live. Declarative knowledge with an attitudinal layer — the kind that benefits from social discussion rather than a proctored exam.
And crucially, the conversation about mental health already happens on social media — loudly, constantly, and often wrong. Delivering assessment there isn't a gimmick. It meets the misinformation on its home turf.
PICKING THE PLATFORM (AND ANSWERING MY OWN PRIVACY QUESTION)
My notes this week were mostly questions: How can I deliver a mini-course on LinkedIn? Would responses be public?
Here's my LinkedIn sketch: a poll as the pre-test (Which of these is a warning sign of depression?), a carousel or short article as the content, a second poll as the post-test, spaced a couple of days apart. Polls are frictionless — one tap — which quiets my time-on-post worry. And LinkedIn fits workplace mental health literacy specifically: managers learning to spot burnout, HR folks learning what to say when someone isn't okay.
But then, privacy. On LinkedIn, the poll author can see exactly who voted for what. For a quiz about mental health, that's a real problem — nobody wants their boss watching them miss a question about depression. Instagram flips this: the Stories quiz sticker shows responses only to the account owner, gives instant right-or-wrong feedback, and evaporates in twenty-four hours. Low stakes by design. And for deeper, discussion-based assessment — peer feedback on scenario responses, say — a private Facebook or Discord group restores the dialogic layer Alonzo et al. (2023) describe, without the public stage.
So the answer to my own question — would responses be public? — turns out to be: only if you choose the wrong tool. Platform selection is a privacy decision, not just a reach decision.
SELF-ASSESSMENT IN THE SCROLL
The self-assessment piece might be the most natural fit of all. A quiz sticker is a self-check the learner grades in the moment — tap, see the answer, feel the tiny sting or the tiny victory, move on. A "how many of these did you know?" checklist carousel invites a private inventory. A story prompt like "What would you say to a friend who told you they weren't okay?" is self-assessment disguised as reflection. Alonzo et al. (2023) found that assessment through social media gives learners a greater sense of participation and ownership — and self-assessment is ownership in its purest form. Nobody's watching. The learner is both examinee and examiner.
MESURING LEARNING, NOT LIKES
Which brings me to the part where my analytics heart beats a little faster. Engagement metrics are seductive and mostly useless as evidence of learning. A thousand likes tells me the graphic was pretty.
Gruzd, Paulin, and Haythornthwaite (2016) offer a better toolkit: pair content analysis with social network analysis to read the digital trails learners leave behind. Applied to my mental health mini-course, I'd measure four things:
The pre/post delta on identical items — the only number that directly claims learning happened.
Content analysis of comments — are misconceptions shrinking over time, is the vocabulary shifting from "snap out of it" toward "warning signs" and "check in"?
Sentiment trends, since tone matters in this domain.
Network structure — are people starting to answer each other's questions instead of waiting for me? Tools like Netlytic can automate real chunks of this. If a peer-support network forms around the content, that's not a side effect. For mental health, the network might be the outcome.
I'm still not sure a feed can hold a whole course. But it can hold a question, an answer, and a correction — and in a domain where wrong beliefs cost lives, that might be enough.
So I'll put it to you — if a mental health quiz showed up in your feed tomorrow, would you tap it? And would knowing your answer was completely private change your mind?
Comments