Publications de Camille Amoura

Introduction

L’évaluation des risques psychosociaux est devenue une composante essentielle de la santé au travail. Il existe de nombreux outils, mais il n’est pas toujours aisé d’identifier leurs qualités psychométriques et les concepts mesurés. L’objectif de ce travail est de présenter une revue critique des principaux outils destinés à évaluer le stress et/ou les risques psychosociaux disponibles et validés en langue française.

Méthode

La recherche documentaire a été effectuée sur les bases de données Medline et PsycINFO, ainsi qu’une revue systématique des Archives des maladies professionnelles et de l’environnement. Les questionnaires entrant dans le champ des risques psychosociaux, et dont il existe une version française, ont été inclus. Les données psychométriques et les concepts mesurés ont été analysés.

Résultats

Onze questionnaires ont été conservés et classés en trois grandes catégories : (i) les questionnaires portant sur les facteurs environnementaux professionnels (i.e. RPS et/ou ‘stresseurs’), (ii) les questionnaires portant sur le processus de stress (i.e. transaction individuelle et/ou stress perçu), (iii) les questionnaires portant sur les conséquences en termes de santé et/ou de morbidité.

Conclusions

Peu d’études publiées portent sur les qualités psychométriques des versions françaises. Il nous apparaît que seuls quelques questionnaires font l’objet d’un concept théorique solide et d’une validation psychométrique suffisante.

The aim of the present studies was to explore if students attending AC (adapted classes - remedial training, 3 times a week; 7-10 years) have a bad reputation (are considered as less able than RC students – Regular Classes) and if this bad reputation can make them vulnerable to stereotype threat. Study 1 assessed self-perceptions of students from adapted (AC) and regular classes (RC) as well as other-perceptions of RC students and meta-perceptions of AC students. Results showed that RC students perceived themselves more positively than they perceived AC students and that self-perceptions of AC students were less positive than those of RC students. They also showed that the meta-perceptions of AC students were similar to the other-perceptions of RC students. Study 2, ran 15 days after, showed that the performance of AC students was impaired when under stereotype threat. Study 3 replicated Study 2 and showed that the effect of stereotype threat on performance was still there even after controlling for initial self-perceptions. It also showed that self-perceptions of AC and RC students lowered after the threat.

Autonomy-supportive and controlling styles of teaching are usually considered to be the opposite ends of a single continuum. An alternative view, however, is that individuals can perceive both styles simultaneously, which suggests that they are different constructs (Bartholomew, Ntoumanis, Ryan, Bosch, & Thøgersen-Ntoumani, 2011). Using cluster analysis, Study 1 (N = 160) confirmed that both teaching styles were perceived by students. Four clusters appeared depending on the student’s score on the measures of autonomy and controlling styles (high autonomy–high control; low autonomy–low control; high autonomy–low control; low autonomy–high control). Participants in the high autonomy–low control cluster reported the highest self-determined motivation in their studies. Using path analysis and mediational analyses, Study 2 (N = 127) tested the independence of the two styles by studying the process through which they influenced motivation. The results showed that need satisfaction (specifically, the need for autonomy) mediated the path between perceived autonomy-supportive teacher behavior and motivation, and that need-thwarting (specifically, the need for autonomy and relatedness) mediated the path between perceived controlling teacher behavior and self-determined motivation, which in turn predicted academic performance. These results add to the existing literature supporting the independence of the two styles.

We examined the psychometric qualities of the Short Form of the Coopersmith Self-Esteem Inventory (SF-CSEI) in a large sample of French adolescents and young adults. A 25-item French version was administered to 1,362 participants (561 aged below 16 years and 801 aged 16-25 years). Participants also completed other scales to measure construct validity (e.g., Rosenberg Self-Esteem Scale, Hospital Anxiety and Depression Scale, and General Health Questionnaire). Factorial analysis yielded evidence for a structure with three first-order factors for the SF-CSEI: personal, social, and family-derived self-esteem. The internal consistency of the questionnaire's different dimensions was satisfactory (Cronbach's α = .68-.77). Pearson's correlation coefficients showed that the SF-CSEI had moderate to high correlations with convergent measures (r = .19-.73) and constructs related to self-esteem (r = -.23-.65). Psychiatric patients (n = 67) scored significantly lower than a control group. Test-retest reliability was good for some of the factors, especially at 5 weeks and 1 year (r = .29-.79). The French version of the SF-CSEI appears to be a useful instrument, with a cross-culturally stable factorial structure.