1. Objectifs

Intentions :
«The current study aims to determine which strategy for combining informant reports is most accurate in identifying children that require clinical services for behavioral problems.» (p. 58)

2. Méthode

Échantillon/Matériau :
«The participants were drawn from a longitudinal study which begun in 2008 in the province of Quebec, Canada. For the purposes of this study, 744 French-speaking Canadian elementary schoolers, aged 6–10 years old […] at the first time of measurement, were recruited. Of these, 370 received complementary school services for emotional or behavioral problems (referred sample) and 374 did not (non-referred sample).» (p. 60)

Instruments :
Questionnaire

Type de traitement des données :
Analyse statistique

3. Résumé

«On the whole, our results regarding the use of informants or report combination strategies are rather consistent regardless of construct measured, child sex, and cutoff (clinical and at risk) tested. If parents and teachers reports taken separately turn out to be of limited usefulness for identifying children in the referred sample as actually having behavioral problems, combining these two informants to calculate a mean T-score does not seem to contribute much more in terms of discriminant validity.» (p. 65) In addition, this study «revealed a stronger discriminant contribution when parent and teacher reports are combined, and the highest T-score is used. Of the different report combination strategies tested in our study, this is the one that demonstrated the best sensitivity, specificity and correct classification rates. This was true for the two cutoffs studied and for both boys and girls. It was also true for the assessment of conduct problems, oppositional problems and attention/hyperactivity problems alike.» (p. 66)