Examining Measurement Properties of the Revised Preschool Language Assessment for Use With Mandarin-Speaking Children

Pei-Hua Chen,Jen-Tso Fu

DOI: https://doi.org/10.1080/15434303.2018.1529176

IF: 2.143

2018-10-02

Language Assessment Quarterly

Abstract:The Revised Preschool Language Assessment (RPLA) is a standardized measure for examining the language status of and determining potential language difficulties among preschoolers between 3 and 6 years old. To facilitate the applicability of the RPLA for use with Mandarin-speaking children, the present study adopted exploratory factor analysis (EFA) and Rasch analysis to assess the appropriateness of using a single total score to interpret children’s language ability. Mandarin-speaking children (N = 138) with no reported disabilities were recruited. Analyses suggested that the RPLA is a useful instrument for identifying differences in children based on language proficiency; however, analyses of total scores showed that the RPLA was not unidimensional; thus, reporting a single total score is not justified. Subsequently, EFA and Rasch analyses were performed in scales, yielding data reflecting separate unidimensional domains of language comprehension and oral expression. The person–item maps indicated that complexity in comprehension and expression increases and provides support for the validity of scores generated by using the measure. Thus, using separate ability indexes for language comprehension and oral expression rather than using one total score was suggested. Finally, applications involving screening and weighting items based on their difficulty levels in multidimensional language assessments were discussed.

linguistics,psychology, educational