Validation française de l’échelle de perception des comportements d'attachement à l’âge préscolaire: version professeur des écoles

Objective: Even if the attachment theory continues to consolidate its conceptual basis by considering children within their overall life context, few tools are available to assess pre-schoolers’ attachment relationship with non-parental attachment figures and more precisely with kindergarten teachers. The aim of this research is to validate a French version of the Portuguese questionnaire “Percepção de Comportamento de Vinculação–Professor” (PCV-P – Attachment Behaviour Perception Scale – Teachers’ version; Dias et al., 2004) and its adaptation to 3 to 6 year-olds. Given its short form and easy administration and data analysis, this scale is an interesting alternative to traditional tools (the Strange Situation, the Attachment Q-Sort) of attachment assessment. Methods: A back translation was led by two bilingual researchers helped by a psychologist and a kindergarten teacher to adapt the French version. With their clinical experiences in toddlerhood, this translation committee adjusted the items for children from 3 to 6 years old; the original version was created for 6 to 8 year-old-children. Twenty-nine kindergarten teachers from 15 French schools answered the questionnaire by quoting their pupils’ attachment behaviours. The French version of the PCV-P, named the PCA-P, is composed of 28 items. The participant has to assess, on a five point Likert scale, the similarity between attachment behaviour statements and the child's behaviours at preschool. Scores are calculated by summing up the items scores into each subscale and for the total scale. The two dimensions score provides an idea of the child attachment in terms of emotional self-regulation and of the quality of his secure base. Moreover, the total scale provides a global “security” score for the child–teacher relationship. Results: The statistical analyses of the PCA-P were carried out on 152 questionnaires. A factorial analysis confirms the two Portuguese original dimensions: “secure-based behaviour” and “emotional self-regulation” and allows to explain 41.83% of the total variance. The Cronbach Alphas (0.86 and 0.89 for the two dimensions respectively, and 0.89 for the total scale) show a satisfying internal consistency. Conclusion: This study reveals good validity and fidelity indicators for the French version of the PCV-P and allows Francophone researchers to assess more quickly and easily teachers’ perception of their pupils’ attachment security in kindergarten. Given the interest for the child to rely on secondary attachment relationships which offer a secure base to explore the environment, the PCA-P opens crucial research and clinical ways in terms of prevention of learning difficulties and school failure. Further studies should provide more data on the psychometric qualities of the PCA-P, namely its convergent, discriminant or test-retest validity.