Theoretical Background
The Science Behind the Signals
From organizational physics to measurement architecture. Sections 1–9 reflect White Paper v1.0. Sections 11–11.7 reflect the Exit Signal Framework addendum, v1.2.
```
Organizational Physics
Organizational physics is a practical metaphor that treats organizations as dynamic systems influenced by forces rather than static structures. The goal is not to claim literal physical laws apply to people; instead, physics concepts provide intuitive language for understanding why some teams hold together under pressure and others fragment.
Gravity as Organizational Attraction
Organizational gravity describes the invisible pull that keeps people connected. Strong gravity emerges when purpose, leadership trust, role clarity, social connection, and equitable treatment align. Weak gravity appears when uncertainty, conflict, misalignment, favoritism, or unmanaged change increases. Coherence Signals translates these invisible forces into measurable signals that can be monitored over time.
Relational Force
Trust, belonging, and team connection.
Psychological Force
Safety, fairness, and perceived support.
Leadership Force
Consistency, responsiveness, and transparency.
Directional Force
Clarity of goals, priorities, and decisions.
Structural Force
Role definition, workload balance, communication flow.
Equity Force v1.2
Fair distribution of standards, recognition, and opportunity.
The Seven Dimensions
Each dimension is differentially weighted by its temporal position in the destabilization sequence, recovery difficulty, and empirical relationship to departure outcomes. Survey items below reflect the v1.2 instrument, revised for a Grade 8 or lower reading level.
Reverse-scored items (marked R) are inverted before index calculation. They are included for response quality assurance and to detect acquiescence bias.
Scoring and Weighting Model
The Coherence Level (CL) is calculated from the seven core dimensions only. CTI compensation items are excluded. Reverse-scored items are inverted, dimensional means are calculated, and the weighted sum is normalized to a 0–100 scale. The Departure Signal Quotient (DSQ) uses a separate weight set that elevates dimensions known to be systematically suppressed in active pulse surveys.
Dimension Weights — CL / DSQ
Psychological Safety20% / 22%
Leadership Response18% / 21%
Equity & Accountability18% / 20%
Connection16% / 16%
Clarity15% / 13%
Change Stability8% / 5%
Future Confidence5% / 3%
DSQ weights are elevated for Safety, Leadership Response, and Equity to account for social suppression in active surveys. These dimensions emerge more prominently in retrospective exit data where candor risk is absent.
Three Interpretive Rules
Equity–Safety Divergence Flag
When Psychological Safety is low and item 21 (favoritism reverse item) scores neutrally or positively, flag potential suppression. Staff in an unsafe environment will not report favoritism candidly. Neutral scores here warrant a specific interpretive note rather than treatment as absence of the problem.
Change Stability Pattern Rule
Single-period dip → communication audit. Three or more consecutive periods of decline → escalate as a Leadership Response and trust issue, not a change management issue.
Alignment Gap Sensitivity
In clinical and high-acuity settings, staff suppress concerns upward while continuing to report laterally. A sudden widening of the Alignment Gap likely means the problem has been held laterally for some time before becoming visible in aggregate scores.
CTI Rule: Compensation (C1 and C2 mean) is reported as Green (>3.5), Amber (2.5–3.5), or Red (<2.5) — a contextual modifier alongside the Coherence Level, never a scored component. See the CTI tab for the full interpretive matrix.
Exit Signal Framework
Addendum v1.2
The Coherence Level measures organizational conditions in real time, from the perspective of currently embedded employees. Exit survey data has a fundamentally different character: it is retrospective, event-driven, and produced by respondents who are no longer subject to the social pressures that suppress candor in active pulse surveys. Exit data is therefore operated as a parallel signal — reported separately, analytically linked to the Coherence Level, and capable of triggering structured adjustments to dimensional weights under defined conditions.
Important: Exit data must not be folded directly into the Coherence Level. Doing so would distort a measure designed to represent current conditions with retrospective departure attribution data.
Function 1 — Standalone Intelligence
Structured Departure Attribution Profile showing which dimensions were most influential in exit decisions, segmented by role level and tenure band.
Function 2 — Retrospective Validation
Does exit data confirm what the Coherence Level flagged in advance? Quantified through the PAI. Discrepancies identify potential blind spots and inform weight adjustment.
Function 3 — Predictive Model Training
Aggregated exit data linked temporally to pulse history identifies how far in advance the Coherence Level begins showing detectable deterioration, and in which dimensions.
Departure Signal Quotient (DSQ)
The DSQ is a composite score representing the weighted pattern of departure drivers over a rolling measurement period. It is calculated from two data streams per dimension: the Experience Score (how deteriorated the condition was during employment) and the Departure Weight Score (how much that condition contributed to the exit decision). The DSQ measures why people left and whether those reasons were visible in advance — not how many left.
DSQ = Σ (Ei × Di × Wi) / Σ Wi
Formula legend:
Ei = Experience Score (1–5; respondent's rated experience of dimension i during employment) ·
Di = Departure Weight Score (1–5; how much dimension i contributed to the decision to leave) ·
Wi = DSQ dimensional weight (see weighting table) ·
i = index across the seven core dimensions. CTI compensation score (Ec × Dc) is reported separately to preserve action-routing logic: relational and psychological findings route to manager-level response; structural compensation findings route to senior leadership.
Predictive Accuracy Index (PAI)
The PAI quantifies how consistently the Coherence Level flagged deterioration in the same dimensions that exit data later identified as departure drivers. A high PAI confirms model accuracy for the given organizational context; a low PAI triggers dimensional weight review.
PAI ≥ 0.70
Model is functioning as an effective early-warning instrument for this organizational context.
PAI 0.60–0.69
Partially predictive. Review dimensional weights and check for systematic suppression patterns.
PAI < 0.60
Formal weight review triggered. CTI discrepancies (green pulse CTI with high exit compensation attribution) logged as potential indicators of compensation suppression in active surveys.
Optional CL Adjustment Mechanism: If exit attribution data over a 12-month minimum rolling period consistently identifies a dimension as a primary departure driver that the Coherence Level did not flag, an upward weight adjustment is applied in subsequent scoring for the affected unit. This adjustment is bounded, transparent, reversible, and documented in the platform's methodology log. · Exit data retention: 36 months to support longitudinal predictive model training. · Reporting threshold: Minimum 5 exit respondents per rolling period before team-level DSQ is reported. · Attribution model: Respondents provide role level and tenure band only — no further identifying information is collected.
Compensation Threshold Indicator (CTI)
Addendum v1.2
Compensation and benefits are excluded from the seven core Coherence Level dimensions because they are structurally different from the other dimensions. They are largely visible, exogenous to team dynamics, and not primarily mediated by perception in the way that psychological safety or leadership trust is. Including pay as a scored dimension also creates an intervention routing problem: the manager playbook becomes incoherent when one dimension requires a response — market repricing, benefits redesign — that sits entirely outside the team manager's span of control.
At the same time, compensation dissatisfaction is empirically among the most frequently cited departure drivers across sectors. The resolution is to treat compensation as a threshold variable rather than a dimensional signal — detected by a two-item screener (C1 and C2) reported separately from the Coherence Level.
CTI Survey Items
Part B — Compensation and Benefits (reported separately from Coherence Level)
C1I feel fairly compensated for the work I do here.
C2The benefits offered here work for me.
CTI = mean of C1 and C2 scores. Reported as a contextual modifier alongside the Coherence Level dashboard. CTI Amber or Red for three consecutive reporting periods surfaces a structural compensation review recommendation to senior leadership.
CTI Status Thresholds
| CTI Status | Mean Score | Interpretation |
|---|
| Green | > 3.5 | Compensation is not a primary stability risk. |
| Amber | 2.5 – 3.5 | Compensation is a moderate concern. Review internal equity and market positioning. |
| Red | < 2.5 | Compensation is likely a confounding variable. Interpret Coherence Level alongside a structural pay and benefits review. |
CTI–Coherence Level Interpretive Matrix
| Coherence Level | CTI Status | Classification | Leadership Implication |
|---|
| Healthy | Green | Stable | Stable conditions. Manager playbook applies normally. |
| Healthy | Amber or Red | Latent Risk | Pay dissatisfaction is a latent departure risk. A market shift or external offer could trigger exits the Coherence Level would not predict. |
| Declining | Green | Relational / Psychological | Instability is genuinely relational or psychological. Manager playbook applies directly. |
| Declining | Amber or Red | Compounding Risk | Compensation is amplifying relational erosion. Structural intervention required above team-manager level. Highest departure velocity risk. |
Dimension 7: Equity & Accountability
Addendum v1.2
The original six dimensions measure relational, psychological, and structural conditions at the team level. Domain expertise from clinical management settings — inpatient, residential, and outpatient — identified a class of upstream destabilization precursors not captured by the existing model: favoritism, clique behavior among staff that excludes key team members, strong performers not being stretched or given advancement opportunity, and weak performers whose issues go uncorrected.
These dynamics share a common structure — perceived inequity in how standards, recognition, and opportunity are distributed — and they appear earlier in the destabilization sequence than most other signals.
Diagnostic gap addressed: The pre-v1.2 model could not distinguish between "this team has low psychological safety because of a poor manager" and "this team has low psychological safety because clique dynamics are operating and staff have learned that raising concerns is socially risky." Equity and Accountability makes that distinction visible.
Survey Items (v1.2 Instrument)
Dimension 7: Equity and Accountability
19Performance problems on my team are dealt with fairly, no matter who they involve.
20People here get fair chances to grow, be recognized, and move forward.
21 ★Some staff are treated differently based on who they know rather than how they perform. (Reverse scored)
Weighting Position
Coherence Level Weight: 18%
Equal to Leadership Response, reflecting its upstream position in the destabilization sequence.
DSQ Weight: 20% — second highest
Favoritism and inconsistent performance management are systematically underreported in active surveys and emerge more prominently in retrospective exit data where candor risk is absent.
Equity–Safety Suppression Pattern
When Psychological Safety is low and item 21 (favoritism reverse item) scores neutrally or positively, the platform flags this as a potential suppression pattern. Staff who have learned the environment is unsafe are unlikely to report favoritism candidly. A neutral score on item 21 in a low-safety environment warrants a specific interpretive note in the leadership dashboard — not treatment as absence of the problem.
```