Rasch-Based Validation of a Japanese-Language Flow Survey: Comparison of Word Cards and Augmented Reality in Deliberate Vocabulary Study
Adam Dabrowski, Ayako YOKOGAWA
This presentation will cover the adaptation of an existing survey of flow into Japanese and validation efforts after conducting the survey after four different deliberate vocabulary study scenarios: word study while reading, a gamified smartphone quiz application, two-sided word cards, and an augmented reality (AR) task. A total of 677 survey responses were collected from 192 participants from three different sites after each activity. The survey data were subjected to Rasch analyses to gather validity evidence to support the hypothesis that the instrument would be capable of detecting differences in flow-state markers between the word card and the AR scenarios. Though a small number of the participants did not display good fit to the expectations of the Rasch model, the majority of the participants displayed good fit, and the items displayed good to excellent fit. Unidimensionality as well as appropriate person and item separation were observed. When participant logit scores were compared on the basis of vocabulary scenario, a statistically significant difference of increased flow markers was observed favoring the AR scenario as compared to word cards. This presentation showcases the process of developing, piloting, and diagnosing a survey as well as its application and findings.