Using the Multilingual Assessment Instrument for Narratives for Evaluating Language Skills in Mandarin-Speaking Children at Early School Ages: Initial Validity

Purpose: Using a longitudinal design, the present study evaluated the utility of the picture sequences in the Multilingual Assessment Instrument for Narratives (MAIN) for assessing Mandarin-speaking children’s narrative skills from kindergarten through Grade 2 in Taiwan. Method: Participants were 27 children with typical language (TL) and 22 with developmental language disorder (DLD) from Taiwan. They were followed up from kindergarten through Grade 2 in approximately 12-month intervals. At each time point, children were administered a norm-referenced, standardized language test and completed both story retell and story generation tasks based on two picture sequences from MAIN. Performance on the narrative tasks was assessed using measures of macrostructure (i.e., story structure complexity) and microstructure (i.e., number of different words, mean length of communication units [C-units], percent grammatical C-units). Results: All target measures showed significant developmental changes and differences between the TL and DLD groups. Correlations between the story retell and generation tasks on each measure were significant across grade levels. In addition, correlations between the target narrative measures and language scores from the tests were mostly significant in both elicitation methods at kindergarten and Grade 1, although the correlations were less robust at Grade 2. Conclusions: Narrative measures derived from the MAIN picture sequences, using story retell and generation tasks, may effectively reflect language skills of Mandarin-speaking children in Taiwan from kindergarten to Grade 2. However, caution is warranted when interpreting results at Grade 2 due to reduced correlations between narrative measures and standardized test language scores.