Les différences individuelles les plus importantes en ce qui concerne la personnalité peuvent être ramenées à cinq groupes de traits ou cinq grandes dimensions nommés « Big Five » : Extraversion, Agréabilité, « Conscienciosité », Névrosisme et Ouverture. Ces cinq facteurs de la personnalité sont retrouvés dans les auto-évaluations, dans le langage naturel, dans des questionnaires, chez les enfants, les étudiants, les adultes et dans plusieurs langues. L’un des instruments les plus utilisés pour l’évaluation de ces cinq dimensions, le Big Five Inventory (BFI) a été validé en France sous sa forme auto-évaluative. L’objectif de la présente étude est de vérifier les principales propriétés psychométriques de la version française du BFI, le BFI-Fr, lorsqu’il est utilisé dans un contexte d’hétéro-évaluation.

Deux cent cinquante-deux étudiants de troisième année de psychologie ont décrit 405 personnes âgées de 18 à 59ans (dont 199 femmes) à l’aide du BFI-Fr, qui comporte, dans sa version française, 45 items. Tous les items sont additionnés pour obtenir les scores moyens.

Les analyses montrent une structure factorielle identique entre la forme hétéro-évaluative et la forme auto-évaluative du BFI-Fr. Les cinq dimensions ont une cohérence interne satisfaisante et les corrélations test-retest sont également toutes très élevées. Les femmes présentent des scores plus élevés que les hommes sur trois dimensions, Extraversion, Agréabilité et Conscienciosité, et les sujets d’âge mûr ont des scores de Conscienciosité plus élevés que les sujets plus jeunes.

La structure factorielle, la cohérence interne et la stabilité temporelle (fidélité test-retest) des échelles obtenues en hétéro-évaluation correspondent aux cinq grands facteurs de personnalité. Le BFI-Fr est donc un instrument pertinent, valide et fidèle pour l’hétéro-évaluation des cinq dimensions de la personnalité issues de la taxonomie des « Big Five ».

The most significant individual differences concerning personality can be reduced into five broad traits or dimensions generally known as the “Big Five Factors”, which may be characterized as: E for Extraversion, Energy, Enthusiasm; A for Agreeableness, Affection, Altruism; C for Conscientiousness, Constraint, Control; N for Neuroticism, Negative emotionality, Nervousness, and O for Openness, Originality, Open-Mindedness. They can be found in self-assessments, in natural language, in questionnaires, among children, students, adults and in several languages. Moreover, self-assessments and assessments by observers are essential methodological tools for personality psychology. The Big Five Inventory is one of the most used measures of these five factors of personality. It was adapted and validated in France in its self-report form. The objective of this study is to validate the use of the French version of the BFI (BFI-Fr) when subjects are assessed by external informants.

Two hundred and fifty-two third-year French students in psychology used the BFI-Fr to describe one or two people they know well. These informants described 405 subjects aged from 18 to 59 years old (199 women). The 45 items of the French version are short statements describing an aspect of personality from adjectives and traits known to be the prototypical markers of the “Big Five”. All items are summed to obtain five average scores. Women generally scored higher than men in dimensions C, A and N. Just as the original American version, the BFI-Fr has a very fast execution time (less than 10 minutes) with a small number of easily understandable items. In order to verify the factor structure of the peer-report BFI-Fr using the same method which was used for the validation of the self-report form, we conducted a principal component analysis with varimax-normalized rotation.

Analysis yielded the same factor structure which had emerged from the analyses of the self-report BFI-Fr. The internal consistency of all scales was satisfactory, with Cronbach's coefficients ranging from 0.84 to 0.87. The test-retest coefficients ranging from 0.86 to 0.90 showed a good stability through time of all dimensions. All correlations between the components are weak, although several are significant. Overall, regardless of age, women scored higher than men on Extraversion, Agreeableness, and Conscientiousness, and regardless of gender, older subjects scored higher on Conscientiousness than the younger ones.

The factor-structure and psychometric properties of the peer-report BFI-Fr do not differ from those of the previously validated French self-report form. Limitations of this study include the fact that all informants were students in psychology trained in the observation of behaviour, which could limit generalisation to other, less psychologically-oriented informants. But on the other hand, it may be assumed that they are reliable observers. The BFI-Fr seems to be a relevant peer-report instrument for the assessment of personality according to the “Big Five” model of personality.