Module 4 · Feedback and motivation

About 20 minutes · Competency domain: Feedback and motivation · competency model

Objectives

By the end of this module you can:

1. Give specific, process-focused feedback — name what in the student’s work was right, not a verdict on the student (behaviors F1–F2, rubric Domain 4).
2. After a correct answer: confirm briefly, ask one why-question, then offer a similar problem (F3–F4) — the exact chain the Socratic AI Tutor encodes as its “confirm and ask why” and “offer a similar problem” moves.

Scenario hook: twenty “good job”s

Priya’s coordinator observes her session and tallies her praise: “good job” ×14, “nice” ×5, “perfect” ×1. The praise is sincere, warm — and nearly useless. Two students who solved correctly can’t say what they did right. One who earned real praise for a repaired sign error hears the same “good job” as the student who copied a neighbor.

Feedback research says process-level, specific feedback outperforms person-level praise — the same evidence base the Socratic AI Tutor’s rules cite (Hattie & Timperley 2007). In a 30-minute small group, that feedback also is the relationship: it’s how students learn that you actually see their work.

Key idea: three habits

1. Specific feedback names the work. “You moved the −4 to the other side and changed its sign — that’s the step you missed last week” tells the student what to keep doing. “Great job” tells the student nothing except that an adult is pleased. A practical bar: your praise should contain one detail that couldn’t be said to any other student.

2. Confirm, then ask why — once. When a student is right: confirm clearly (“That’s right — x = 5”), ask one why-question (“Why did we subtract before dividing?”), let them answer, then move on. The why is where success turns into understanding; endless questioning after a right answer teaches students that being correct buys more interrogation, which teaches them to hide correctness.

3. Offer a similar problem — don’t assign it. After a solved problem: “New one, same shape: 4x − 3 = 17. Want to try it — I’ll say less this time?” This is the cheapest honest test that the fix transferred, and the offer (not the assignment) keeps ownership with the student.

Normalize errors as you go: “That’s the trap this problem sets — let’s look at it” beats “no, again?” every time. Students who believe errors are information take the risks that produce thinking.

Model: a solved problem, closed properly

Verbatim excerpt from Transcript T1 (fictional, constructed for training) — the end of guided practice: confirm, ask why once, then offer the similar problem:

Luis: …plus 5 both sides? So 2x = 16.

Nina: Say why plus.

Luis: To move the −5… and the sign flips because we’re subtracting minus 5.

Nina: You kept both sides balanced — that’s the move. Now x? [confirm and ask why]

Luis: Divide by 2 — x = 8!

Nina: Check it for us.

Luis: 2(8 − 4) + 3… 8 + 3 = 11. Yes! [the student checks his own work]

Nina: Same shape, new numbers: 3(x + 2) − 5 = 16. Solo, four minutes — I’m not talking. If you finish, check by substituting. [offer a similar problem]

What to notice: the confirmation names the process (“you kept both sides balanced”), the why comes once and then the tutor moves on, and the similar problem arrives as an offer with the tutor going silent — ownership stays with the student. (For the specific-feedback habit — naming the work, not the person — see T1’s opening exchange and annotation 6.)

Practice with feedback

Your turn — upgrade these three. Rewrite each as specific feedback (one detail, process-focused):

1. “Good job!” — after Amara correctly distributes in 3(x + 2) − 4.
2. “You’re so smart.” — after Luis solves 4x = 12 without hints.
3. “Perfect!” — after Mei fixes her own sign error on the second try.

Strong rewrites:

1. “You distributed to both terms — the +2 didn’t escape this time.”
2. “You divided every term by 4 in one move — no hints from me on that one.”
3. “You checked your signs before I said anything — that’s exactly the habit from last session.” Why: each names the behavior that produced the success, so the student knows what to repeat. “Smart” teaches students to protect an identity instead of practicing a process — and it collapses the first time they struggle.

In your next session, try…

1. Give every student one piece of specific feedback tied to their recorded work (tracker).
2. After each correct answer: confirm → one why → offer a similar problem. Three chains in one session is a good session.
3. Catch one error early and say “that’s the trap this problem sets” — out loud, warmly, before anyone else notices.

Reflection

In the tutor self-reflection form: Whose success did I praise most this week? What did I actually name? Which student never heard their own work described?

Check yourself

• A student solves fast and asks for harder work. Confirm-and-why, similar problem — or jump to new content? (Offer a similar problem first: one small variation tests whether the method — not just this answer — transferred. Then escalate.)
• A student is visibly deflated after a third error. What sequence? (Normalize briefly, then the worked example after repeated failure (L4): work a similar problem fully, then hand back the original. Comfort without a plan just postpones the frustration.)