Abstract
This study examines the validity of score-based interpretation of the Story Retelling Speaking Test (SRST) in comparison with the Versant(Pearson Education, Inc., 2008) and Standard Speaking Test (SST; ALC Press, Inc., 2010). In total, 64 participants took the three tests; their speaking functions, scores, and utterances were analyzed to probe the shared and varied aspects of the tests. The results showed that the SRST elicited more functions than the Versant but fewer than the SST, that it was moderately related to the latter two tests, and that it more successfully discriminated among a group of beginner and intermediate level learners. Additionally, the results suggested that (a) the tasks and speaking functions and (b) the aspects emphasized while ratingmay explain the differencesin the test scores for the three tests.Based on the results, comparative advantages of each test were summarized, which may be useful for selecting appropriate speaking tests according to assessment purposes and situations.
本研究では、Story Retelling Speaking Test (SRST) を、Versant (Pearson Education, Inc., 2008) と Standard Speaking Test (SST; ALC Press, Inc., 2010)と比較し、SRSTの得点に基づく解釈の妥当性を吟味する。64名の受験者が3つのテストを受けた。テストの共通点と相違点を調べるために、その言語機能と得点、発話が分析された。その結果、SRSTはVersantより多いがSSTよりは少ない言語機能を引き出すこと、SRSTは他の2つのテストと中程度の相関があり、初級者・中級者グループで弁別力を持つこと、(a)タスクと言語機能と(b)評価時に強調する要素の相違により、テスト得点の違いが説明されうることが分かった。結果に基づき、評価の目的と状況に沿って適切なスピーキングテストを選ぶ際に有益となる、各テストの相対的な利点を示した。
