According to a large-scale survey administered by the Ministry of Education, Culture, Sports, Science and Technology, Japan, in 2021, interview tests wherein students speak with a teacher or an assistant language teacher are mainly used in Japanese junior high schools to assess interactional competence in speaking English. However, there has been a lack of investigation regarding the types of utterances elicited by students during interview tests conducted by teachers. This study aimed to fill this gap by involving a teacher who interviewed 41 junior high school students, allowing for a holistic and analytical examination of the dialogues during the interviews. The analysis of these student-teacher dialogues revealed that the teacher actively assisted students in continuing the dialogue by offering new topics, developing them, and asking topic-related questions to continue the dialogue. The authors argue that these findings indicate that the interviews demonstrated the teacher's interactional competence more than that of the students, making it difficult to assess students' interactional competence through teacherconducted interviews. Therefore, alternative speaking-test formats are needed to assess the students' interactional competence in English.

Measures of intercultural understanding are part of a growing body of language testing scholarship focused around defining L2 proficiency as a sociocognitive process, that is, as the interplay between learner-internal language knowledge and learner-external cultural factors. The performance assessment described in this paper used an analytic rubric to assess the sociocognitive variable of intercultural understanding through an interview task. To demonstrate the relationship between the descriptors on the rubric and authentic task performance, we analyzed transcripts of the interview task itself, along with official feedback reports about the test from raters and administrators. To create a probabilistic model of rater agreement, we used Many-Facet Rasch Analysis to evaluate the extent to which a high score on this task indicates an examinee's ability to function well in intercultural situations, and a low score indicates the opposite. The results provide some evidence for a unidimensional construct related to intercultural understanding that can be measured using an interview task.

This study proposes a better usage of sentence recognition tasks to measure the effects of retelling on the textbase and situation-model representations of Japanese learners of English as a foreign language. The effects of retelling were analyzed from two perspectives: Does retelling reinforce the memory of textual information and help construct a more elaborate situation model? Does retelling improve sensitivity to identifying any contradictory information at the textbase or situation-model levels? The participants’ memories were analyzed using the recognition confidence ratings of the test items (paraphrases, inferences, and distractors). A total of 25 Japanese students from the National Institute of Technology participated in the experiment. All participants read one experimental text and answered sentence recognition tasks. Subsequently, they read the other text and retold the story in Japanese before taking on the sentence recognition task. The learners’ mental representations were analyzed in two steps. First, recognition confidence ratings for paraphrases and inferences were analyzed as the strengths of the textbase and situation-model representations, respectively. Second, the difference values were analyzed as the sensitivity to identifying contradictory information at each level of understanding. The results suggested that retelling significantly strengthened the memory of information at the textbase level and led to a more elaborate situation model. The results further demonstrated that retelling significantly improved the sensitivity to detecting contradictory information at the situation-model level. These findings underscore the efficacy of using a two-step sentence recognition task to precisely elucidate the effects of retelling on reading comprehension.