Teacher-Student Rapport Dynamics

Teacher Action

Auto-teach (runs on its own)

Class Setup

Class size 14

Rapport decay rate k 0.15/s

Vicarious effect strength 0.35

Selected Student

Click a student in the 3D view to select them.

Classroom State

Avg. rapport

0.0%

Avg. engagement

50.0%

Interactions

Disengaged

How it works

Each student i carries a rapport score R_i ∈ [-1, 1] that drives their visible engagement. Every teacher action is an event with a target value v (praise = +1, neutral = 0, correction = -0.7) aimed at one student, and it updates the whole room at once — directly for the target, weaker for everyone else through observational (vicarious) learning:

for each student i:
  w_i = 1                       if i is the target
  w_i = s · exp(-d(i,target)/2) otherwise   (s = vicarious strength, d = seat distance)
  R_i ← R_i + η · w_i · (v − R_i)      (η = 0.5 learning rate)

between events, rapport relaxes toward neutral:
  dR_i/dt = −k · R_i            (k = decay rate slider)

Engagement_i = clamp(0.5 + 0.5·R_i, 0, 1)

Direct reinforcement — the targeted student's rapport moves toward the action's value fastest.
Vicarious effect — classmates update too, more if they sit closer to the interaction (Bandura's observational learning, and Wubbels' findings that teacher-student interpersonal behaviour is witnessed, not private).
Decay — with no reinforcement, rapport drifts back to neutral, so a relationship needs upkeep, not a single good day.
Click a student to select them, then use Praise / Neutral / Correct to target that student directly, or let Auto-teach run the class on its own.

Click a student to select · drag to orbit

High rapport Neutral Low rapport

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