Science

Fondement scientifique

Nos évaluations sont basées sur des modèles psychométriques validés et une recherche continue pour garantir fiabilité et validité.

Recherche scientifique
α > 0,8
Fiabilité interne

Nos évaluations respectent les plus hauts standards de qualité psychométrique.

α > 0,8
Fiabilité interne
> 0,85
Stabilité temporelle
Content
Validité de contenu
Construct
Validité de construit

Modèles validés

Why4® s'appuie sur des modèles reconnus mondialement : DISC, Spranger, l'Intelligence Émotionnelle (Goleman), l'Intelligence Culturelle (Earley & Ang, 2003) et le modèle cognitif CHC.

  • DISC : 4 dimensions comportementales éprouvées
  • Spranger : 6 motivations profondes validées
  • Intelligence Émotionnelle : modèle de Goleman (5 piliers)
  • Intelligence Culturelle : modèle CQ (Earley & Ang, 2003) — 4 dimensions
  • Cognitif : modèle Cattell-Horn-Carroll (CHC)

Validation psychométrique

Nous maintenons des standards élevés de qualité psychométrique avec des études de fiabilité et de validité régulières.

  • Coefficients de fiabilité (Cronbach's α) supérieurs à 0.80
  • Validité de contenu et de construit démontrée
  • Études de validation continue sur des populations diverses

Recherche & Innovation

Notre équipe R&D travaille en permanence à améliorer nos outils et à développer de nouvelles approches.

  • Intégration de l'IA pour des insights plus approfondis
  • Recherche appliquée sur l'intelligence émotionnelle
  • Architecture auto-évolutive — chaque évaluation affine le modèle prédictif
  • Conformité aux standards internationaux : ISO 10667, APA, EFPA
Whey — Intelligence artificielle propriétaire

Un Moteur d'Intelligence Auto-Évolutif

Whey ne se contente pas d'analyser le marché du travail — il apprend en permanence de chaque évaluation complétée. Son architecture bicéphale — LLM propriétaire + World Model prédictif — forme une boucle fermée d'amélioration continue.

Le pipeline Whey

De la donnée brute au score psychométrique calibré

  1. 1
    Ingestion massive
    1,5M résultats/mois · 504 flux RSS · LinkedIn.
  2. 2
    Analyse LLM
    Mistral extrait les dimensions comportementales implicites : Dominance, Influence, Stabilité, Conformité, EQ, CQ.
  3. 3
    Contextualisation régionale
    Le World Model applique les pondérations culturelles CH/EU/GCC — invariance de mesure entre marchés.
  4. 4
    Calibration psychométrique
    IRT + DIF · Cronbach α > 0,85 · test-retest > 0,88 · alignement ISO 10667 et standards APA.
  5. 5
    Boucle de rétroaction
    Chaque évaluation complétée réalimente le World Model via meta-learning. Le système s'améliore en continu.
Le cœur du système

La boucle d'auto-évolution Whey

Chaque évaluation complétée, chaque offre d'emploi ingérée, chaque profil analysé réalimente le World Model. Le système ne stagne jamais — il converge vers une précision croissante.

Collecte continue
1,5M de signaux/mois — offres, profils, publications — ingérés et normalisés automatiquement.
Analyse LLM
Le LLM extrait les dimensions comportementales implicites et génère des scores bruts par poste et région.
Correction World Model
Le World Model prédictif contraint, valide et recalibre les scores selon les normes culturelles régionales.
Meta-learning
Les distributions de réponses réelles réalimentent le World Model via few-shot learning — le système s'améliore à chaque cycle.
Ce que le système garantit
> 0,85
Cronbach α
> 0,88
Test-retest
Mensuel
Cycle de mise à jour
Références académiques

Bibliographie scientifique

1
Fondement théorique DISC
Marston, W. M. (1928)
Emotions of Normal People
Harcourt, Brace & Company
2
Fondement théorique Why6®
Spranger, E. (1928)
Types of Men
Steinkopf (Université de Berlin)
3
Validation instrument DISC
Geier, J. G. (1979)
D.I.S.C. Personal Profile System: Manual
Performax Systems International
4
Big Five - NIH
Costa, P. T., & McCrae, R. R. (1992)
Revised NEO Personality Inventory (NEO PI-R)
Psychological Assessment Resources
5
Meta-analyse - 162 études
Barrick, M. R., & Mount, M. K. (1991)
The Big Five personality dimensions and job performance
Personnel Psychology, 44(1), 1–26
6
Université de Floride
Judge, T. A., Heller, D., & Mount, M. K. (2002)
Five-factor model of personality and job satisfaction
Journal of Applied Psychology, 87(3), 530–541
7
Harvard Business Review
Goleman, D. (1998)
Working with Emotional Intelligence
Bantam Books / Harvard Business Review Press
8
Yale / UNH
Mayer, J. D., Salovey, P., & Caruso, D. R. (2004)
Emotional intelligence: Theory, findings, and implications
Psychological Inquiry, 15(3), 197–215
9
Université du Texas
Bar-On, R. (2006)
The Bar-On model of emotional-social intelligence (ESI)
Psicothema, 18(Suppl.), 13–25
10
Australie
Schutte, N. S., Malouff, J. M., et al. (2009)
Exploring the validity of a measure of trait emotional intelligence
Journal of Personality Assessment, 91(5), 444–450
11
Harvard - Psycho-lexical
Allport, G. W., & Odbert, H. S. (1936)
Trait-names: A psycho-lexical study
Psychological Monographs, 47(1), i–171
12
16PF - Illinois
Cattell, R. B. (1943)
The description of personality: Basic traits resolved into clusters
Journal of Abnormal and Social Psychology, 38(4), 476–506
13
Université Tulane
Hogan, R., & Holland, B. (2003)
Using theory to evaluate personality and job-performance relations
Journal of Applied Psychology, 88(1), 100–112
14
Yale - Concept original
Salovey, P., & Mayer, J. D. (1990)
Emotional intelligence
Imagination, Cognition and Personality, 9(3), 185–211
15
Michigan / Western Ontario
Tett, R. P., Jackson, D. N., & Rothstein, M. (1991)
Personality measures as predictors of job performance
Personnel Psychology, 44(4), 703–742
16
Norme internationale
ISO 10667 (2020)
Assessment service delivery — Procedures and methods
International Organization for Standardization (ISO)
17
Guidelines internationales
International Test Commission (ITC) (2001)
International Guidelines for Test Use
ITC Guidelines (Revised 2013)
18
Standards APA
American Psychological Association (APA) (2014)
Standards for Educational and Psychological Testing
APA / AERA / NCME
19
Standards européennes
European Federation of Psychologists Associations (EFPA) (2013)
EFPA Review Model for the Description and Evaluation of Psychological Tests
European Psychologist, 18(2), 87–99
20
Standards UK
British Psychological Society (BPS) (2017)
Testing and Assessment: Guidance on Test Use
BPS Psychological Testing Centre
21
Source fondatrice CQ
Earley, P. C., & Ang, S. (2003)
Cultural Intelligence: Individual Interactions Across Cultures
Stanford University Press
22
Modèle psychométrique CQ — 4 dimensions
Ang, S., Van Dyne, L., et al. (2007)
Cultural Intelligence: Its Measurement and Effects
Management and Organization Review, 3(3), 335–371
23
CQS — Cultural Intelligence Scale
Van Dyne, L., Ang, S., & Koh, C. (2008)
Development and Validation of the CQS
Handbook of Cultural Intelligence. M. E. Sharpe
24
Sous-dimensions du modèle CQ
Van Dyne, L., Ang, S., et al. (2012)
Sub-Dimensions of the Four Factor Model of Cultural Intelligence
Social and Personality Psychology Compass, 6(4), 295–313
25
Synthèse académique CQ
Van Dyne, L., Ang, S., & Nielsen, T. M. (2016)
Cultural Intelligence
Oxford Bibliographies in Management
26
CQ — Domaine et évaluation
Thomas, D. C., et al. (2008)
Cultural Intelligence: Domain and Assessment
International Journal of Cross Cultural Management, 8(2), 123–143
27
CQ vs QI/QE — Efficacité managériale
Rockstuhl, T., et al. (2011)
Beyond General Intelligence (IQ) and Emotional Intelligence (EQ)
Journal of Social Issues, 67(4), 825–840
28
CQ et résultats de négociation
Imai, L., & Gelfand, M. J. (2010)
The Culturally Intelligent Negotiator
Organizational Behavior and Human Decision Processes, 112(2), 83–98
29
CQ — Personnalité et métacognition
Chao, M. M., Takeuchi, R., & Foo, M. D. (2017)
Deciphering Cultural Intelligence
Academy of Management Proceedings, 2017(1)
30
CQ dans l’industrie hôtelière
Chen, A. S., Lin, Y. C., & Sawangpattanakul, A. (2011)
Cultural Intelligence and Performance in the International Hospitality Industry
International Journal of Contemporary Hospitality Management, 23(1), 84–98
Demander une démo