Conceptualisation et validation factorielle d’un questionnaire mesurant le profil émotionnel : Emotional State Questionnaire (ESQ) - 17/02/08
pages | 10 |
Iconographies | 4 |
Vidéos | 0 |
Autres | 0 |
L’objectif de cette étude est d’évaluer la validité d’un nouvel autoquestionnaire, l’Emotional State Questionnaire (ESQ). Cet instrument a pour but de mesurer le profil émotionnel, c’est-à-dire le fonctionnement émotionnel d’une personne et non la réactivité émotionnelle face à des stimuli. Il a été construit à partir des 6 émotions de base décrites par Ekman et selon 4 dimensions : la reconnaissance, le ressenti, l’expression et le contexte social et relationnel. À l’issue de la phase conceptuelle, 33 items ont été générés. Tous les items sont quantifiés selon 5 modalités, de pas du tout à toujours (extrêmement ou très fortement). La validation de structure interne (analyse d’item, étude de la convergence item-dimension détermination de la structure factorielle, évaluation de la cohérence interne) a été réalisée chez 218 sujets issus de la population générale et ne présentant pas de dépression et/ou d’anxiété pathologique. Les propriétés psychométriques obtenues sont satisfaisantes compte tenu du concept mesuré. L’acceptabilité est bonne avec un taux de données manquantes inférieur à 5 %. L’analyse factorielle permet d’identifier les quatre dimensions postulées qui restituent 42 % de la variance totale. Les trois premières dimensions présentent une bonne cohérence interne (α > 0,70). La cohérence interne de la dimension contexte social est plus faible (α = 0,58) car elle rassemble l’aspect social impliqué dans la reconnaissance, l’expression et le ressenti. Elles présentent des scores moyens répartis de façon homogène et se situant autour de la valeur centrale.
Conceptualisation and validation of the « Emotional State Questionnaire (ESQ) » : evaluation of an emotional profile |
Objectives –The objective of this study is to evaluate the validity of a new self questionnaire : the « ESQ » (Emotional State Questionnaire). Background – This novel instrument possesses a number of original attributes : first of all, it is designed to assess a general emotional profile, in opposition to other similar scales which can only be applied to the emotional reactions provoked by specific stimuli. Secondly, this scale is composed of several emotional dimensions. The ESQ has been constructed according to four components : recognition, expression, internal emotional experience and social context. The first three dimensions were selected because of their wide use through behavioral experiments. Indeed, contrary to most scales used in this field, which only assess the emotional experience, we wanted to propose an instrument also able to assess the subject’s impression of his own capacities to encode and decode emotions. We hypothesized that these three dimensions could not be dissociated from a fourth dimension, the social context, which therefore also figures in this scale. The emotions explored were the five fundamental emotions indicated by Izard (fear, happiness, sadness, disgust and surprise) to which we added a neutral feeling that we considered as a basic emotion. Study design –To establish this instrument, a first conceptual phase was conducted by a group of experts. These experts all worked in the psychological field. They proposed the scale on the base of their clinical experience and after study of the literature. The scale was then validated in a population of 218 healthy volunteers, aged between 15 and 88 years. Subjects were not included if they presented depression (score above 16 in the Beck Depression Scale) or pathological anxiety (score above 5 in the Spieberger State Anxiety Inventory). The psychometric characteristics tested were : the item analysis, the item-dimension correlation, the factor analysis and the internal consistency reliability. Results –The population studied was equally distributed according to gender (sex ratio : 0.97), the mean age was of 36.2 2 ± 16.1 years. Acceptability was good with less than 5 % of data missing. The analysis of items revealed no floor or ceiling effect and a low correlation between items. Item-dimension correlation ranged from 0.23 to 0.62, with most scores above 0.4. The items were always better correlated to their dimension than to other dimensions, except for one item. The 4 dimensions (recognition, expression, internal emotional experience and social context of emotions) explained 42 % of the total variance. Finally, the scale showed good internal consistency with Cronbach coefficients, equal or above 0.84 for the total score, the recognition and the expression dimensions. This coefficient reached 0.77 for the feeling dimension but only 0.58 for the social context dimension. Conclusion –All together, results showed satisfactory characteristics in regard of the complexity of the notion measured. However, an important drawback is the lack of an external instrument to assess convergent validity. This instrument can be of great interest in the emotional characterization of healthy volunteers. Moreover, if validated in psychiatric populations, this scale could be most useful in psychopathological assessment and also in comparison with behavioral evaluations of emotion.
Mots clés :
Analyse factorielle
,
Émotions
,
Psychométrie
,
Questionnaire; Validation.
Keywords: Emotion , Factor analysis , Psychometric properties , Questionnaire , Validation.
Plan
© 2007 Elsevier Masson SAS. Tous droits réservés.
Vol 33 - N° 2
P. 169-178 - avril 2007 Retour au numéroBienvenue sur EM-consulte, la référence des professionnels de santé.
L’accès au texte intégral de cet article nécessite un abonnement.
Bienvenue sur EM-consulte, la référence des professionnels de santé.
L’achat d’article à l’unité est indisponible à l’heure actuelle.
Déjà abonné à cette revue ?