VIC Framework - Variability-Integration Cleft

This subject integrates better than predicted by their pediatric variability (C_subject > +1 sigma).

This subject integrates less well than predicted by their pediatric variability (C_subject < -1 sigma).

This subject follows the predicted trajectory (C_subject within ±1 sigma).

SYNTHETIC DATA - R9 simulator anchors (N=500/phenomenon, seed 42). Not a diagnosis. Not patient triage. No real patient data.

Click a point to view its C_subject, or pick one below.

I = 0.5 functional + 0.3 masking + 0.2 allostatic

C2: Window closes before adult measurement

C4: No cohort spans the window

0.25 V + 0.25 I + 0.50 C

0.40 cohort + 0.40 instruments + 0.20 access

VIC is NOT a diagnosis and is NOT a patient-level triage tool.

ALL current anchors are SYNTHETIC (simulator), not empirical patient data.

VIC is a research framework for prioritising research phenomena, not for evaluating individuals.

Version v1.0.0 - preliminary specification. Parameters will change with empirical data.

A multidisciplinary scoring tool for critical developmental windows

Introduction: What is the VIC Framework?

The VIC Framework (Variability-Integration Cleft) is a research specification that names and quantifies a precise methodological problem: for many developmental phenomena, the critical plasticity window closes before adult instruments apply, and no individual cohort follows the same subjects across that window. VIC names this configuration and scores individual traversal.

The tool was developed by Guillaume Desvaux and published under MIT and CC-BY-4.0 license. The current version is v1.0.0 - a preliminary specification. All current numerical anchors are synthetic (R9 simulator, N=500 per phenomenon, seed 42) and not empirical patient data.

The cleft: no cohort follows the same subjects from childhood to adulthood.

A persistent methodological problem runs through developmental neuroscience: pediatric studies document early variability, adult studies measure integration costs, but no cohort connects the same individuals across the critical window. Critical windows - those periods of highly elevated neural plasticity - close before standardized adult clinical instruments become applicable.

VIC names this configuration: a pediatric study A, an adult study B, non-overlapping time windows, and the structural impossibility for a single cohort to cover both. The cleft (C) is the quantification of what cannot be predicted about adult integration from pediatric variability.

This is not a statistical power problem. It is an ontological constraint: the same subjects cannot be measured before and after a window that closes, under identical conditions. VIC formalizes this observation and proposes a scoring framework to prioritize research phenomena.

The VIC triangle: V (pediatric variability), I (composite adult integration cost), C (residual cleft).

V measures the dispersion of the pediatric developmental trajectory, normalized to [0, 1]. A high V indicates strong individual variability during the critical window; a low V indicates homogeneous development. V is constructed from pediatric study data (segment A).

I - Adult integration cost (composite)

I is a composite index measuring the adult functional integration cost. It is calculated as I = 0.5 * I_functional + 0.3 * I_masking + 0.2 * I_allostatic, with each sub-axis normalized to [0, 1]. The default weights (0.5, 0.3, 0.2) can be adjusted per disciplinary context. I is constructed from adult study data (segment B).

C is the measure of the residual unpredictability of I from V. At cohort level: C_cohort = Var(I - I_predicted) / Var(I) = 1 - R^2. A high C_cohort means a large share of adult integration variance is not explained by pediatric variability - that is the cleft. At subject level: the standardized residual C_subject = (I_observed - I_predicted) / sigma_residual indicates whether an individual integrates better (C > +1 sigma) or less well (C < -1 sigma) than predicted by their childhood variability.

The four VIC qualification conditions. All four must be met.

A developmental phenomenon meets VIC qualification if and only if all four conditions are simultaneously met. A single unmet condition disqualifies the phenomenon.

C1: A critical window exists - plasticity is significantly elevated during a bounded period of pediatric development.

C2: The window closes before adult measurement - standardized clinical instruments are only applicable after the window closes.

C3: Both segments are documented separately - pediatric studies (segment A) and adult studies (segment B) exist for this phenomenon, but not a single cohort covering both.

C4: No cohort spans the window - no single longitudinal cohort follows the same subjects from the pediatric window through adulthood for this phenomenon.

This filter is intentionally strict. It eliminates phenomena for which complete longitudinal data already exist (C4 not met) or for which the critical window is not established (C1). It targets cases where the gap is structural, not surmountable by simply increasing statistical power.

Scoring flow: Severity and Tractability feed Priority.

For a cohort, C_cohort = 1 - R^2 from the I sim V regression. It is the fraction of adult integration variance that cannot be predicted from pediatric variability. A high value (e.g., 0.421 for visual_asd) indicates a large cleft; a low value (e.g., 0.082 for attachment_bpd) indicates that pediatric variability predicts adult integration well.

VIC computes three domain scores to help prioritize research efforts. These scores apply to the research domain (phenomenon), not to an individual patient:

Severity = 0.25 * V_disp + 0.25 * I_disp + 0.50 * C - weights axis dispersion and the cleft, with the cleft at double weight.

Tractability = 0.40 * cohort\_feasibility + 0.40 * instrument\_availability + 0.20 * accessibility - measures the practical ease of conducting the missing longitudinal study.

Priority = Severity / max(Tractability, 0.05) - severe and easy-to-study phenomena get highest priority. The floor of 0.05 avoids division by zero.

VIC has been applied to seven candidate developmental phenomena, spanning neuroscience, psychiatry, and developmental psychology. visual_asd is the worked example. The six scored phenomena have calibrated synthetic anchors (R9 sim, N=500, seed 42). The hormonal_identity phenomenon is deferred due to instrument heterogeneity.

The following table reproduces the published synthetic metarubric anchors. These values were generated by the R9 simulator with N=500 subjects per phenomenon and random seed 42. They are NOT empirical patient data. They serve as calibration anchors for the synthetic demonstrator.

SYNTHETIC NOTE: All numerical values in the tables above are synthetic simulator anchors (R9, N=500, seed 42), not empirical measurements of real patients.

The interactive demonstrator below replays the simulator anchors - not patient data. It deterministically recomputes, from the published metarubric parameters (beta_1, sigma_residual, seed 42), an illustrative V vs I scatter for each of the six scored phenomena. The regression line corresponds to the published beta_1 slope. Selecting a subject displays their standardized residual C_subject.

VIC is a research framework under development. Its v1.0.0 version is a preliminary specification. The CLINICAL_NOTICE.md from the source repository states the following limits, reproduced here verbatim:

The composite I weights (0.5/0.3/0.2) are illustrative defaults, not empirically validated. The qualification thresholds for the four conditions (C1-C4) require disciplinary operationalization - VIC provides the framework, not the thresholds. Data availability for Tractability sub-axes (cohort feasibility, instrument availability, accessibility) requires expert domain assessment. No phenomenon yet has empirical anchors; all current parameters are synthetic.

The path toward empirical anchors requires longitudinal studies that cover the critical window to adulthood transition - precisely the studies VIC aims to help prioritize. This bootstrapping is inherent to the specification and acknowledged as such in the documentation.