Multi-Channel Affective Measurement

Description

Measures emotional response by triangulating self-report, physiological signals and behaviour in the same session, exposing mismatches between what users say and what they feel.

Emotion is not a single signal. Self-report captures what people are willing and able to articulate; physiology (electrodermal activity, heart rate variability, pupil dilation) captures arousal without valence; behaviour (facial expression, hesitation, posture, interaction patterns) captures expression shaped by social context. Picard's affective computing framework treats each channel as a partial, noisy measurement of an underlying affective state. The rigour comes from declaring channels and hypotheses before collection, synchronising all channels to interface events, and treating disagreement between channels as data rather than error: a participant who says a flow was 'fine' while skin conductance spikes at a fee disclosure is telling you something no single channel could. Post-hoc selection of whichever channel supports a story is the characteristic failure the protocol is designed to prevent.

Attributes

Loss type
Siloed
Stage
Evaluation
Data
Mixed
Effort
Weeks

Steps

  1. Declare all channels before data collection. Post-hoc triangulation is selection bias.
  2. Collect all channels simultaneously or within the same session.
  3. Implicit vs. explicit mismatch analysis: Participant says 'it was fine' but shows elevated GSR and long hesitation times. The mismatch IS the finding — report it, do not average it away.

Sources

  • Affective Computing (Rosalind Picard)
  • Methods Atlas — affective measurement (Gopinath)
  • Affective Computing (Picard)
  • The Inquiry Series, Ch 37 — Picard's Affective Framework — Multi-Channel Emotion Measurement

Also known as

  • Affective triangulation
  • Picard's multi-channel emotion measurement
  • Self-report, physiological and behavioural triangulation

Categories

  • Analytics & Behaviour
  • Neuro, Biometrics & Affect