Pedagogical Feedback Simulation

Feedback Delivery

Feedback frequency 6 / min

Feedback delay 3.0 s

Specificity / quality 60%

Praise ↔ Correction 40% corrective

Class State

Avg. mastery

Avg. motivation

50%

Feedback efficacy

At mastery (≥80%)

0 / 30

Feedback rounds given

How it works

Each of 30 simulated students carries a skill level S ∈ [0,1] (bar height) and a motivation level M ∈ [0,1] (bar color, red→green). A feedback round fires at the chosen frequency, then lands on the class after the chosen delay:

efficacy   = specificity · e^(-delay / 8)
ΔS_i       = η · efficacy · M_i · (1 − S_i) · r_i
ΔM_i       = specificity · (1 − 1.6·|balance − 0.4|) − 0.15

between rounds:  S_i -= decay · S_i · dt   (forgetting)
                 M_i -= (M_i − 0.5) · relax · dt

Feedback frequency — how often a round is delivered; more rounds mean more compounding growth but each is smaller per student.
Feedback delay — time between attempt and feedback. Efficacy decays exponentially with delay, matching Hattie & Timperley's finding that immediate feedback is far more effective than delayed feedback.
Specificity / quality — how actionable the feedback is ("check your carry digit" vs "good job"); scales both the skill gain and the motivation response.
Praise ↔ Correction balance — motivation peaks near a mostly-praise, some-correction mix (≈40% corrective); all-praise feels empty, all-correction erodes motivation, and low motivation r_i throttles how much a student actually absorbs.
r_i is a fixed per-student receptivity (0.6–1.4) so the class responds unevenly, as real classrooms do.

Real-world relevance: this is the mechanism behind why frequent, immediate, specific, balanced feedback outperforms rare, delayed, vague feedback in every major education meta-analysis — and why harsh, all-corrective feedback can raise skill briefly while quietly collapsing motivation.

High motivation Low motivation Class average (line)

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.