Search for a command to run...
Résumé : Cette recherche analyse la structure factorielle de l’engagement au travail via une mesure à 18 items couvrant les dimensions cognitives, affective et comportementale. Les données ont été recueillies auprès de 202 professionnels RH via une échelle de type Likert à 5 points. Respectant les recommandations méthodologiques pour l’exploration de construits latents, une analyse factorielle exploratoire a été réalisée (extraction en axes principaux, rotation oblique Oblimin) pour identifier une structuration empirique parcimonieuse et guider la purification de l’instrument (Fabrigar et al., 1999). Les résultats plaident pour une solution à trois facteurs et révèlent des regroupements d’items partiellement divergents de l’organisation théorique attendue, suggérant à la fois la constitution d’un noyau d’engagement plus intégré et une différenciation inégale de ses facettes. Ces résultats conduisent à poursuivre par une validation confirmatoire (AFC/CFA) comparant des modèles alternatifs, ainsi qu’à vérifier la validité discriminante, en ligne avec les conceptualisations de l’engagement comme investissement multidimensionnel du soi au travail. Abstract : This study examines the factorial structure of work engagement using an 18-item instrument encompassing cognitive, emotional, and behavioral dimensions. Data were collected from 202 HR professionals using a five-point Likert scale. In line with methodological recommendations for the exploration of latent constructs, an exploratory factor analysis was conducted (principal axis factoring with oblique Oblimin rotation) to identify a parsimonious empirical structure and to guide scale purification (Fabrigar et al., 1999). The findings support a three-factor solution and reveal item groupings that partially diverge from the expected theoretical organization, suggesting the presence of a more integrated core of engagement alongside an uneven differentiation of its facets. These results call for subsequent confirmatory validation (CFA) comparing competing measurement models (correlated factors versus hierarchical or bifactor models), as well as for the assessment of discriminant validity, in line with conceptualizations of engagement as a multidimensional investment of the self-in-work roles.