Self-assessment of your university teaching practice with AI

Context

A university lecturer or professor who wants to examine a specific aspect of their own teaching without waiting for the formal student evaluation cycle or a visit from their institution's teaching and learning center. Use it after a class that raised doubts, before rethinking a unit, or as part of a personal professional development exercise. This does not generate evidence for external teaching reviews or accreditation processes — it is a tool for reflection.

Originally written for Chilean classrooms. References to Chilean institutional bodies (university consortium, teaching and learning centers, national accreditation agency) have been adapted for an international audience.

Paste first

Before opening the model, prepare:

• Course name, program, semester, and institution
• Course profile: group size, format (in-person, hybrid, online), nature (lecture, practical, workshop)
• A free-form description of what concerns you about your current teaching — it might be participation, comprehension, content coverage, assessment, or classroom climate
• A description of a recent class (the most recent one you remember clearly, or the one that concerned you most), without student names

Privacy rule: Do not include identifiable student data (names, individual grades). Describe situations: "two students fell asleep" is useful context; naming them is not necessary and should not go into an external AI model.

Act as a university teaching consultant with experience supporting faculty across a range of institutions — from large research universities to teaching-focused regional colleges. Your role is not to evaluate me: you have no access to my student course evaluations or my grades. Your role is to be my critical thinking partner so I can identify patterns in my own teaching and propose concrete actions to improve it.

**Academic context:**
- Course: {{course_name}}
- Program and semester: {{program}}, {{semester}}
- Institution: {{institution}}
- General course profile: {{course_description}} (estimated size, format — in-person / hybrid / online, whether it is a lecture, practical, or workshop)

**What concerns me about my current teaching:**
{{free_description_of_concern}} (e.g., low participation, high failure rates, difficulty running active classes, tension between content coverage and deep understanding, not enough time for feedback)

**Description of a recent class (no student names):**
{{description_of_recent_class}}

I want you to respond in three parts:

**Part 1 — Pattern diagnosis:** Based on what I described, identify between two and four pedagogical patterns that seem to recur in my teaching. For each pattern, name it precisely, describe how it manifests in what I shared, and explain what possible consequences it may have on student learning.

**Part 2 — Unresolved tensions:** Identify one or two pedagogical tensions I might not be seeing — things that seem contradictory or that imply a choice I have not made explicit. Frame them as questions, not diagnoses.

**Part 3 — Prioritized improvement plan:** Propose three concrete actions I can implement in the next two weeks without redesigning the whole course. Each action must specify: (a) exactly what I do, (b) when I do it (which moment in the class or in the course cycle), and (c) how I will know whether it had an effect.

Use the language of a peer, not an evaluator. When in doubt between stating and asking, ask.

Expected output

Example (condensed):

Pattern diagnosis (excerpt)

Pattern 1 — Lecture as the dominant structure: 20 minutes of slides before the
case study positions students as receivers. By the time they are asked to
participate, they have been in passive mode for half an hour. The silence is not
because the group is passive — the structure trains them to be.

Pattern 2 — Question without scaffolding: "What type of organizational structure
does this company have?" requires students to recall the concept, apply it to a
new case, and have the confidence to say it aloud — three simultaneous steps
with no support.

Unresolved tension: Are you prioritizing covering the unit or verifying it was
understood? If you left class feeling most students had not grasped the concept,
what did advancing the syllabus accomplish?

Improvement plan — action 1 (Think-Pair-Share): Before asking the group, give
90 seconds for each student to write their answer on paper, then one minute in
pairs. You will know it is working if more than one person responds
spontaneously during the whole-group discussion.

Watch out for

• The model does not know your institution, your department, or your history with this group. Its patterns are hypotheses based only on what you described — use them as a starting point, not as a definitive diagnosis.
• If the model proposes large structural changes (redesigning the syllabus, implementing a completely new methodology), ask for alternatives that fit within a single 90-minute session using the resources you have today.
• Do not use what you generate here as input for institutional student evaluations, your institution's accreditation process, or academic promotion reviews. This reflection is personal — it does not have the methodological validity those processes require.

Suggested next step

If the diagnosis feels too generic: "The patterns you identified could apply to any university class. Go back to the description I gave you and tell me what specific detail from that class led you to identify each pattern."

If you want to go deeper on one action: "Action 2 seems relevant to me. Give me three variations of that same action adapted to a classroom with 60 students and no technology available."

If you want to connect this to institutional standards: "Review the tensions you identified and tell me which one could relate to the teaching quality criteria in my institution's accreditation process. Do not evaluate me — just flag them so I can decide whether they are relevant."