This study examines the relationship between how learners of Japanese as a second language perceive the learning of kanji (i.e., the logographic characters shared with Chinese) and their ability to learn novel kanji words using morphological and contextual information. Eighty college students learning Japanese as a foreign language completed a 60-item kanji questionnaire, a 75-item kanji test, and a 30-item reading comprehension test. Results indicated modest but statistically significant correlations between the belief variables and the kanji ability measures, with reading proficiency factored out. Regression analyses revealed that although reading proficiency accounted for a large portion of variance of the participants' performance on the kanji test, belief in the effectiveness of metacognitive strategies accounted for 14-16% of the variance of success in morphological analysis. The results suggest that (a) students' task-specific beliefs have a significant impact on their achievement on a given task and that (b) metacognitive awareness significantly affects how a learner handles a challenging learning task.

This study explored the pedagogical benefits of flipped online kanji (i.e., Chinese characters used in Japanese) instruction integrated into college‐level introductory and intermediate Japanese language courses. Using a quasi‐experimental mixed design, the investigation looked at the effects of two instructional approaches: (1) a flip approach in which students learned new characters using online materials prior to class, and (2) the conventional (nonflip) approach in which in‐class lectures were followed by printed exercises. During the 2014–2015 academic year, a total of 46 students who were enrolled in introductory or intermediate courses learned 9 to 12 sets of 10 new kanji characters per semester through one lesson per week. Half of the lessons were taught in the flip format, and the other half were taught in the nonflip format. The results showed that introductory students scored better on postlesson tests in the flip condition than in the nonflip condition only for fall 2014 but not for spring 2015, whereas no statistically significant difference was observed for intermediate students. However, the students in both courses reported that online exercises helped them improve their kanji knowledge. Consequently, a substantial portion of kanji instruction can be shifted to self‐paced, semi‐structured online learning with no negative impact on instructional quality.