| Popis: |
Purpose: Interviews to obtain statements in judicial procedures need to be adapted to the witnesses’ abilities to testify. Moreover, knowing the cognitive abilities involved in testifying provides relevant criteria to assess statement credibility. As age or intelligence quotient is not enough to estimate these capabilities, an instrument to evaluate witnesses’ specific abilities to testify is needed. The present paper validates CAPALIST, a procedure that considers relevant capabilities when assessing the testimony given by children.Methods: This study analyzed, by means of an invariant measurement approach (Rasch model), four scales included in CAPALIST: language, memory, contextual information, and social thinking. In addition, gender and age differences were analyzed in 83 children [45 males and 38 females; Mage = 4.3 years, SD = 0.74, range (3.06–5.11)] from three courses in early childhood education.Results: The four scales do not severely violate the requirements of the model. The principal component analysis of the residuals indicates that the four scales are one dimensional and that the assumption of local independence was not violated. Differential item functioning of the scales associated with gender was not detected. A significant effect of the school year was obtained, with an increase in ability in successive courses. The percentage of children who presented severe misfit responses with the model was low. In addition, the number of items with a severe misfit was also low.Conclusion: An acceptable performance of CAPALIST is demonstrated for most of the scales, although items with a severe misfit must be replaced, and more difficult items have to be included in some scales of the revised version of the instrument. CAPALIST is a promising procedure to assess the abilities of children to testify in order to adapt interviews and to evaluate their statements correctly. |
