Using Think-Aloud Protocols in Second Language Research

Concurrent verbal reports, or think-alouds, are a data elicitation technique used in a variety of fields, including Second Language Acquisition (SLA), to gain insight into the cognitive processes involved while performing a task (Ericsson & Simon, 1993; Bowles, 2010). These verbal reports allow researchers to observe internal processes by providing data on the mental states heeded by individuals when carrying out a task and derive information about the relevant mental processes. Although think alouds offer the advantage of making an individual’s ‘inner speech’ audible, they have been criticized on the grounds that they may be reactive or non-veridical (e.g., Ellis, 2001; Jourdenais, 2001; Nisbett & Wilson, 1977). In spite of this criticism, however, many researchers agree that verbal reports can provide direct evidence about cognitive processes that would otherwise remain unavailable (e.g. Hayes & Flower, 1983; Polio, 2012). In SLA research in particular, it is difficult “to determine the reasons behind a learner’s target language use” (Bowles, 2010: 8) which practitioners usually infer from the learner’s output, including the types of errors she makes. Data collected through verbal protocols can, for example, provide insight into a learner’s awareness of L2 forms, her ability to detect errors and perceive feedback, as well as into how her interlanguage changes and develops. In light of this, the use of think-alouds as a research tool in both qualitative and quantitative studies will be the focus of my paper. I will begin with an overview of verbal reports, followed by a review of studies which have employed think-alouds in SLA and conclude with some guidelines for using think-alouds.