Error analysis of expressive analogy task in Spanish-English bilingual school age children with and without specific language impairment

Purpose: The relational shift hypothesis (RSH) states that, as children age, the way in which they interpret analogies shifts from a focus on object similarities to relational aspects of objects. This study investigated the validity of the RSH by describing the error patterns of typically developing (TD), low normal (LN), and language impaired (LI) bilingual school-age children when completing an expressive analogy task in A:B::C:D format (e.g. good:bad::happy:_____) in English and Spanish. Method: Participants included a total of 49 Spanish-English bilingual children between the ages of 7;4 and 8; 9 (mean = 8; 1). Ten children were identified as LI, ten scored in the LN range, and 29 were TD. Children were administered English and Spanish versions of the item twice, initially during the second grade and once again approximately one year later. Responses were recorded verbatim and coded as correct (C), thematic/category error (THEM/CAT), wrong object, correct relationship error (WO-CR), unrelated error (UNREL), or repetition/no response (REP/NR). Results: A repeated measures ANOVA was used to compare children’s analogy scores by time, ability, and language. Results demonstrated significant differences for ability. Four chi square tests investigated the error patterns of TD, LN, and LI bilingual children in English and Spanish. We compared responses provided children by response type (C, THEM/CAT, WO-CR, UNREL, or REP/NR). Results from the Spanish analogical reasoning task indicated a decrease in THEM/CAT with age for the LN and TD children. Results from the English analogical reasoning task were inconsistent. Conclusions: Results provide partial support for the RSH in LN and TD children, but not in children with LI. This difference in error patterns may provide insight into the validity of the RSH in bilingual children with specific language impairment and typically developing second language learners.