test-takers’ perceptions of paired speaking tests and the role of interlocutor variables in pairing

As a mixed-method study, this study investigated efl test-takers’ perceptions of paired speaking tests and the role of interlocutor variables (i.e., peer interlocutor’s acquaintanceship, gender, and proficiency level) in pairing in a cambridge english language assessment open center. the participants (n = 148), having experienced sitting either live paired speaking tests of cambridge esol (group p, n = 53) or only ielts solo interview, (group s, n = 48) or both (group p-s, n = 47), completed a validated likert-scale questionnaire. also, 63 of the participants attended, on a voluntary basis, a semistructured interview. findings indicated group p test-takers had significantly more positive perceptions of paired/group speaking tests than their counterparts in group p-s, who, in turn, had significantly more positive perceptions of pairing than group s participants. interview results revealed that the majority of the participants preferred friend, same-proficiency-level, and same-gender peer interlocutors. the (culture-specific) reasons for the findings are discussed in detail.