Running an Anonymous AMA in Class (Without Chaos)
Anonymous AMA in class, without chaos. When AMA beats lecture, how to set expectations, moderation patterns, and three formats that work.
Hushwork Team
An AMA in class works when it's the right tool for the right session. An hour of "ask me anything" can be the most useful class of the term, or it can be a fishing expedition that frustrates everyone. The difference is preparation, expectations, and moderation.
When AMA beats lecture
Three moments where AMA outperforms regular teaching:
- End of a unit. Students have absorbed enough to know what they don't know. An AMA surfaces the gaps you were going to find on the exam anyway, with time to fix them.
- Right before a major project or exam. Anxiety is high; the questions are concrete. An hour of AMA replaces a queue of office-hours visits.
- First class of the semester. Students don't know each other, don't know you, and have wildly different expectations. An anonymous AMA lets them ask the things they wouldn't out loud yet.
Don't run an AMA every week. The point is its scarcity; if every Tuesday is "ask me anything," nobody saves up the good questions.
How to set expectations
Before the AMA, tell students three things:
- Anonymous means anonymous. "Your name is not attached to the question. I won't know who asked. The point is so you ask things you wouldn't normally raise."
- Scope. "You can ask about course content, the project, the exam, the field, my career, anything. I might decline a question, but I'll explain why."
- What you'll do with bad questions. "If a question is mean or off-topic, I'll skip it. We're spending the time on the real ones."
Saying this out loud at the start changes the question quality dramatically. Students who weren't sure if anonymous really meant anonymous see the pattern from the first answered question and trust the channel.
Three formats that work
Format 1: rolling AMA (45 minutes)
Open the AnswerLink dashboard on the projector. Questions arrive live. You see them appear, pick one, answer, pick another. Hush AI groups duplicate questions and ranks by topic so you don't waste time on three versions of the same query.
Best for: end-of-unit catch-ups, exam prep.
Format 2: pre-collected AMA (30 minutes prep, 60 minutes class)
Open AnswerLink a week in advance. Students drop questions whenever they think of them. The day of class, you've had time to look at the cluster Hush AI surfaced and prepare answers for the deeper ones.
Best for: career-and-field AMAs, complicated topics, last-class-of-the-term reflections.
Format 3: split AMA (40 minutes class)
Run for 20 minutes live, then take a 5-minute break. During the break, look at unanswered questions Hush AI has clustered for you. Come back, address the clusters in the second 15 minutes.
Best for: first-time AMAs where you don't know what to expect.
Live moderation patterns
You will get questions you don't want to answer. Three patterns for handling them:
The off-topic question ("what's your favourite restaurant?")
You can answer it briefly if it lightens the room, or skip with a one-liner ("save that one for office hours"). Either is fine. Don't make it the third one you answer.
The personal question ("what's your salary?")
Answer the version of the question that's actually about teaching or career, not the literal one. "I won't share my salary, but here's what someone in this kind of role typically earns and what the trajectory looks like." That's the question they meant.
The hostile question ("why is this course so badly designed?")
Don't engage with the framing. Re-ask: "What's the specific thing that's not working for you?" If the answer surfaces, you have a real piece of feedback. If it doesn't, you've shown the room that hostility doesn't get a stage.
Hush AI flags hostility patterns before they reach your dashboard, so most of these get filtered already.
What students will ask if it's actually anonymous
Surprising patterns that show up:
- The basic question they were too embarrassed to ask in week one ("I still don't fully understand what a derivative is"). Answer it. Half the room is grateful.
- The career question ("how did you actually decide on this field?"). Real curiosity. Answer with specifics, not platitudes.
- The "why are we doing this" question about a specific assignment. The honest version of "what's the point." Sometimes the answer is that the assignment is worth keeping; sometimes the question is teaching you that it isn't.
- The "what would you do differently" question about your own career. Students are figuring out their own path; your honest answer is more useful than abstract advice.
What to avoid
- Don't read every question out loud. Some are duplicates, some are off-topic. Hush AI groups them; address the cluster, not each instance.
- Don't quote the question wording verbatim if it's identifying. Paraphrase to the gist.
- Don't promise to answer every question. You won't have time for all of them. Pick the high-signal ones.
- Don't end the AMA without a follow-up. "Drop questions you didn't get to ask on AnswerLink and I'll answer in the next class."
After the AMA
The questions you didn't get to are signal. Put them in your prep notes for next year's version of this lecture. The same questions will come up; you'll be ready for them.
The questions that surprised you most are the most useful. They tell you about a gap between how you think the course is landing and how it actually is.
A simple template
Sign in to Hushwork. Open the AnswerLink dashboard on the projector. Open the class with "drop your questions". Close with "we'll keep AnswerLink open for the rest of the week."
Get started
Free for any class size, no student accounts needed.
Related reading:
