摘要 英語學習者之口說能力於近年內廣受重視,因而發展具信效度的口語測驗因而更具重要性。圖片與文本廣泛使用於考試之中,然而對於此兩種提示類型對受試者口語能力影響的研究卻付之闕如。因此,口語評量題型與提示類型對於受試者口語能力的影響之重要性不言可喻。本研究探討不同型態之提示方式對學習第二外語大學生之口語能力測驗之影響。該研究選擇兩種常見的提示題型:文本提示與圖片提示。52 名台灣北部為英文主修的大學生參與此次實驗。每位參與者皆接受此兩種題型之口語測驗。考試結果經分析,以探究提示方式對於大學生口語能力的影響以及考試者對提示題型的分析歷程。兩位評分者評量受試者之口語內容後分別給予整體分數與個別分數,包含正確性、流暢性、內容豐富度。此外,研究者訪談了七位不同程度表現的受試者,以期分析出不同程度受試者之分析歷程。成對樣本T 檢定之結果顯示兩種提示題型並不會對受試者的口語整體或是個別表現有明顯影響。然而大多數受訪之受試者亦表示他們較喜愛圖片式的提示方式,原因在於受試者能夠藉由圖片導向的提示產出綜合性的資訊,而面對文本導向的提示受試者往往只能產出片段性的資訊,且受試者認為從圖片導向的提示考題中擷取非語文類的內容較為容易。更多對於未來設計口語題型之研究建議亦附在討論部分。
ABSTRACT English learners’ speaking ability has been increasingly valued in EFL context in recent years, and thus, the tasks of developing valid and reliable assessment of EFL learners’ speaking ability has become more important. Pictures and texts have been widely used in tests. However, few studies have reported the influence of the types of prompts on test takers’ oral performance and their test –taking process. Therefore, it is critical to examine the characteristics of task prompts that may bring impact on EFL students’ oral performance in speaking assessments. This study investigated the impact of different types of prompts in assessing EFL college students’ oral proficiency. Two commonly used prompt types in second language oral assessments (i.e., text-cued and picture-cued prompts) were selected. Fifty-two English-majored college students in the northern Taiwan participated in this study. Each took two oral tests –one picture-cued and one text-cued prompts. Test outcomes were analyzed to gauge the empirical impact of prompt type on EFL college students’ oral proficiency performance and test takers’ processing of different types of oral prompts. Two raters evaluated test takers’ oral responses and assigned holistic and analytical scores in terms of three aspects, accuracy, fluency and content complexity. In addition, seven test takers of various proficiency levels were interviewed to collect in-depth information regarding the test-taking process of test takers when they responded to different prompt types. Results of the paired-sample T tests showed that these two types of prompts did not exert statistically significant influence on EFL college students’ oral performance holistically or analytically. However, the majority of interviewed test takers reported that they preferred picture-cued prompts because it was easier for them to generate integrative ideas and when responding to text-cued prompts, they tended to reproduce fragmental ideas which were directly taken from text-cued prompts. They also claimed that it was easier to retrieve non-linguistic information from picture-cued prompts. Finally, study implications and suggestions for designing oral assessment tasks were discussed.