A case for the use of the ability-in language user-in context orientation in game-based assessment
- Alexandra Lay, Micheline Chalhoub-Deville
OBJECTIVE The purpose of this study was to test whether and how performance on a digital game-based social skills assessment tool, "Zoo U" (3C Institute, Durham, NC), differed for children in the United States and Japan across six core social skills. MATERIALS AND METHODS "Zoo U" was administered to 497 third and fourth grade children from the United States and Japan (46 percent Japanese) by teachers and researchers, respectively. U.S. children received the original version of "Zoo U," and Japanese children received a fully translated Japanese version of the program. Scoring of each of the six social skills is built into the "Zoo U" software, with specific scoring algorithms for each grade level that provide both a continuous scale score and cutoffs for three distinct performance categories: high, average, and low. RESULTS A multivariate analysis of variance (MANOVA) was conducted to assess differences by cultural group on each of the six continuous social skill scores, controlling for gender and grade level. Results revealed significant differences on four of the six skills in expected directions. Chi-squared and odds ratios analyses were then conducted on the assignment of children into each of the performance categories by cultural group, revealing additional nuance to the cultural differences identified in the MANOVA consistent with existing literature. CONCLUSIONS We were able to replicate known cultural differences between U.S. and Japanese children with a simple direct translation of a Web-based social skills assessment game, "Zoo U." Our results provide preliminary support for the potential of game-based assessment methods to provide efficient and valid social skill assessments to children around the world.