How Situation Models Explain Difficulty in Japanese EFL Reading

Evidence From Eye Tracking

Abstract

Successful reading comprehension requires constructing coherent mental representations of texts (i.e., situation models). However, the mechanism of maintaining coherence in EFL reading has not been fully clarified. Therefore, the Tsukuba Reading Research Group explored this issue using eye-tracking methodology. The participants read narratives that included an inconsistency between a character description (e.g., a vegetarian) and his/her subsequent action (e.g., ordering a cheeseburger). In Experiment 1, the maintenance of local coherence was examined by inserting a single intervening sentence between these conflicting statements. Results showed that local inconsistency increased the initial processing time of the characters’ action and caused more look-backs to the earlier character description. In Experiment 2, one or four intervening sentences were inserted to examine the maintenance of local and global coherence. Results indicated that local inconsistency induced more look-backs whereas global inconsistency failed to do so, indicating that the participants experienced difficulty in maintaining global coherence. In Experiment 3, the participants were instructed to mentally visualize the situation when reading narratives including global inconsistency. Results demonstrated that the instruction helped maintain global coherence, as evident from the processing time of the character’s action. These results were discussed based on the cognitive processes of situation-model construction.