The present study investigated students’ preferences for the types of tasks used to assess English speaking performance. It further examined whether students’ task type preferences affected their perceptions of test effectiveness. One hundred eighty-two high school students responded to a self-report questionnaire. A series of frequency analysis and paired samples t-tests were used for the analysis. The results showed that students’ most preferred task types and their least preferred ones overlapped with each other, suggesting that the task types of English-speaking performance tests used in schools are limited. The four key reasons determining students’ task type preferences were identified, including task difficulty, emotional comfort, practical value, and interest. In addition, the results indicated that students’ task type preferences could affect their perceptions of task effectiveness. Overall, the results suggest the need for developing more varied task types for English-speaking performance tests as well as helping students become familiar with English speaking performance tasks. Pedagogical implications were discussed along with study limitations.

This study investigated the effects of pairing based on English proficiency and gender on high school students’ speaking task performance. A total of 16 high school students - 4 female advanced, 4 female intermediate, 4 male advanced, and 4 male intermediate learners - performed two information gap tasks spotting differences between two pictures, once with a same-level learner and once with a different-level learner. Their performance was analyzed in terms of degree of task completion, amount of utterances and fluency. The results showed (a) advanced level learners performed the task more accurately, more fluently, and in higher length when they were paired with advanced level learners than paired with lower level learners, although the differences were not significant; (b) For intermediate level learners, male students demonstrated a better performance when paired with higher level learners, but female students performed better when paired with the same level learners. The findings are discussed with regard to more feasible and more effective ways of grouping for pair work in high school English classes.