Dissertation/Thesis Abstract

Assessing interactional competence in second language paired speaking tasks
by Wang, Linxiao, Ph.D., Northern Arizona University, 2015, 236; 3713923
Abstract (Summary)

The last decade has witnessed growing awareness and use of paired speaking tasks in the field of second language (L2) assessment, resulting in calls for more studies on interactional competence (Ducasse & Brown, 2009; May, 2010; Taylor & Wigglesworth, 2009). This dissertation study aimed to account for the nature of peer-peer interaction in L2 paired speaking tasks through the perspectives of second language acquisition and task-based language assessment. This has been accomplished by addressing the following issues: first, understanding the construct of interactional competence at both macro- (i.e., overall interaction quality determined by degree of collaboration and task completion) and micro-levels (i.e., particular features in interactions); second, evaluating rating scales for interactional competence in terms of reliability and validity; third, examining the extent to which the distribution of interaction features predicted interaction scores; and lastly, investigating how task type affected interaction performance regarding interaction features, interaction patterns, and interaction scores. In total, 70 language learners in an English for Academic Purposes (EAP) program were paired and responded to four 2.5-minute interaction tasks. All interactions were scored on overall interaction quality by four raters using a new scale developed by the researcher and two existing scales. Then each individual interlocutor’s performance was coded for interaction features ranging from interactive listening to interactional management features. The data obtained were analyzed to answer the following questions: (1) Were hypothesized relations among interaction features supported empirically? (2) Was the new scale a reliable and valid measure? and (3) Did interaction features account for variance in interactional competence scores? Results helped refine understanding of interactional competence. First, the results of confirmatory factor analysis did not support the original model of interactive listening and interactional management features. Instead, interaction performance was better characterized by different communication functions of argument, discussion, support, and connection. Second, compared with the two existing scales, the new scale was a more reliable and valid measure. Third, features of topic connection, turn interruption, and turn overlapping were important indicators to predict scores generated by the new scale. Findings have implications for second language acquisition, L2 speaking assessment, and instruction. First, interactional competence could be operationalized in aspects of interaction features and patterns. Second, speaking assessment should include paired speaking tasks to reflect test-takers’ interactional speaking abilities. In addition, interaction rating scales should measure levels of interactions and task completeness status. Rater calibration is needed to ensure that consistent judgments can be delivered. Last, teachers can direct students’ attention to different interaction patterns and specific features, which characterize effective conversations. In summary, findings provide further understanding of interactional competence and offer insights into how to measure interaction competence in language classrooms.

Indexing (document details)
Advisor: Jamieson, Joan, Kang, Okim
Commitee: Kim, YouJin, McGroarty, Mary
School: Northern Arizona University
Department: English
School Location: United States -- Arizona
Source: DAI-A 76/12(E), Dissertation Abstracts International
Subjects: Linguistics, English as a Second Language, Language
Keywords: Interaction features, Interactional competence, Peer-peer interaction, Rating scales, Task-based language assessment
Publication Number: 3713923
ISBN: 9781321907643