Emotional regulation signals — response latency patterns, communication consistency, escalation or de-escalation trajectories — are among the most predictive behavioral dimensions. They are also among the most dangerous to score without careful governance.

The Engines

  • Emotional Regulation Engine — Scores the behavioral indicators of emotional regulation capacity: consistency of tone, escalation rate, recovery time after conflict signals, and communication pattern variance.
  • Dysregulation Detection Engine — Detects patterns consistent with emotional dysregulation without attaching diagnostic labels. Outputs: "elevated stress indicators", not "borderline personality disorder".
  • Regulation-Decision Correlation Engine — Measures the correlation between regulation indicators and decision quality in the same actor over time. High-regulation actors make more consistent decisions.

Code Walkthrough

// Safe emotional regulation scoring
function scoreEmotionalRegulation(signals) {
  const indicators = {
    toneConsistency:  computeToneVariance(signals.messages),
    escalationRate:   computeEscalationRate(signals.timeline),
    recoveryTime:     computeRecoveryAfterConflict(signals.conflicts),
    patternVariance:  computeCommunicationVariance(signals.patterns),
  };

  const rawScore = weightedAverage(indicators, REGULATION_WEIGHTS);

  // Governance: translate internal scores to safe labels BEFORE returning
  return governanceWrapper.wrap({
    score:  rawScore,
    label:  rawScore < 0.4 ? "elevated-stress-indicators" : "stable-regulation-indicators",
    // NEVER: "emotionally_dysregulated", "mental_health_concern", "unstable"
  });
}

What to Watch For

  • Never output diagnostic language. The governance wrapper enforces this — but do not rely on the wrapper alone. The engine itself should never generate clinical terms.
  • Require human review for any score below 0.35. Low regulation scores in legal or employment contexts are high-consequence outputs.