3. Investigating the empirical link between task-based interaction and acquisition

A linguistic perspective on communicative language teaching

“…The influential Interaction Hypothesis of Michael Long (1981Long ( , 1983Long ( , 1996 argues that language development depends on learner interaction including input, output and negotiation of meaning. This claim finds support in numerous empirical studies, with Keck et al (2006) providing a useful meta-analysis.…”

The Quantity and Quality of Language Practice in Typical Interactive Pair/Group Tasks

“…This interest has since been extended to an examination of the impact of different task types and task features on acquisition (e.g., Robinson, 2001;Skehan & Foster, 1997, resulting in a sufficient number of studies to permit a meta-analysis of the impact of task features on the acquisition of grammar and lexis (Keck, Iberri-Shea, Tracy-Ventura, & Wa-Mbaleka, 2006). Keck et al (2006) conclude, however, that the pedagogical implications of their analysis of the 14 studies 1 reviewed are limited by at least three factors. The first is that most of the research has focused on a restricted range of task types (just two of the six potential types in the framework used: information gap and jigsaw tasks).…”

“…The tasks consisted of interviews, games, descriptions, comparisons, giving and following directions, discussions, consensus, predictions, and writing tongue twisters. Note that none resemble the picture differences, dictogloss, or jigsaw tasks that have figured in previous research on tasks and interaction (Keck et al, 2006; see also Mackey, 1999;Pica, 1987;Pica, Kang, & Sauro, 2006;Swain & Lapkin, 2000.…”

Scope and research methodologies

“…However, the meta-analysis by Keck et al (2006) concludes that in total, experimental groups do seem to outperform comparison groups. One problem with generalizing from a body of studies like this is that the wealth of studies also uses a wealth of experimental designs from more controlled to relatively free tasks on pairs or groups of learners performing on a range of task types.…”

“…Yet, it has also been noted that negotiation may be more relevant to interaction between nonnative and other nonnative speakers, since for many learners this is more likely than interaction with native speakers. In a meta-analysis by Keck et al (2006), however, 85% of studies still involved native -nonnative speaker interaction. Another limitation from the point of view of generalizability is that the majority of research tends to be conducted in university settings as this is where researchers have most immediate access to learners.…”