Assessing Speaking Proficiency: A Narrative Review of Speaking Assessment Research Within the Argument-Based Validation Framework

摘要

This paper provides a narrative review of empirical research on the assessment of speaking proficiency published in selected journals in the field of language assessment. A total of 104 published articles on speaking assessment were collected and systematically analysed within an argument-based validation framework (Chapelle, Enright, & Jamieson, 2008). We examined how the published research is represented in the six inferences of this framework, the topics that were covered by each article, and the research methods that were employed in collecting the backings to support the assumptions underlying each inference. Our analysis results revealed that: a) most of the collected articles could be categorized into the three inferences of evaluation, generalization, and explanation; b) the topics most frequently explored by speaking assessment researchers included the constructs of speaking ability, rater effects, and factors that affect speaking performance, among others; c) quantitative methods were more frequently employed to interrogate the inferences of evaluation and generalization whereas qualitative methods were more frequently utilized to investigate the explanation inference. The paper concludes with a discussion of the implications of this study in relation to gaining a more nuanced understanding of task- or domain-specific speaking abilities, understanding speaking assessment in classroom contexts, and strengthening the interfaces between speaking assessment, and teaching and learning practices.