Workshop survey — pre and post

Same items before and after. Five confidence items (5-point Likert) + two open questions. Anonymous; takes about 5 minutes. Collect the pre-survey via the calendar-invite link only — it is not run in-session (five Likert items plus two open questions do not fit in the opening’s spare minutes, and the opening poll already competes for attention). Collect the post-survey during minutes 82–90 while pairs write their commitments (QR code on screen). Run it as a form (LMS/Forms) or print both halves of this page.

Part 1 — Confidence items (5 = strongly agree … 1 = strongly disagree)

Rate yourself right now:

1. I can look at one of my own assessments and judge which rubric criteria generative AI satisfies easily.
2. I can set an explicit AI-use level for each stage of an assignment, and explain the choice to students.
3. I know at least three practical redesign moves for AI-era assessment, with their workload costs.
4. I could redesign one of my own assignments this month so the grade certifies something AI cannot produce alone.
5. I can explain to a student why our AI-use rules are what they are, without hiding behind a detector.

Post-survey wording change (only): “I can” → “After this workshop, I can”. Everything else is identical — that is what makes the pre/post comparison meaningful.

Part 2 — Open questions

1. What is the one assignment you most want to redesign, and what is stopping you? (Pre-survey. The post-survey asks: “What will you do differently with that assignment in the next two weeks?”)
2. What would make this redesignable for you — time, policy coverage, department support, examples from your discipline?

Data handling

• Anonymous by design: no names, no emails, no free-text asking for identifying details.
• Participants who miss the pre-survey (calendar-invite only) are excluded from the pre/post means rather than surveyed mid-session — say so in the summary.
• Aggregate reporting only (pre/post means per item, themes from open text).
• If you collect canvases with workshop permission, store them separately from survey data.
• Share the summary back with participants — the loop closes when they see what changed.

What this survey can and cannot show

• Can show: self-reported confidence change after one session; barriers named by participants.
• Cannot show: whether redesigns actually happen (that’s the two-week commitment follow-up), or student learning effects. Say this in your summary — it models the honesty the kit teaches.