Investigating The Empirical Link Between Task-based Interaction And L2 Acquisition:A Meta-analysis

Since Long first put forward the Interaction Hypothesis(1981),a large number of studies that focused on interaction have emerged(e.g.,Gass&Alvarez Torres 2005;Yang&Lyster 2010;Rassaei 2014;Li,Zhu&Ellis 2016).Among numerous research issues,the relationship between task-based interaction and second language acquisition is a hot one.Some studies(e.g.,Mackey 1999;Yang&Lyster 2010;Van de Guchte et al.2015)held the point that task-based interaction can facilitate the acquisition of specific linguistic features empirically,but there were also some studies(e.g.,Loschky 1994;Gass&Alvarez Torres 2005;Erlam&Loewen 2010)drawing different conclusions.This paper is intended to do a meta-analysis study to synthesize the findings of all experimental studies published between 2004 and 2017 on the relationship between interaction and second language acquisition.Participants of this study were 11 primary studies published between 2004 and 2017,and these unique sample studies were ultimately identified by an exhausted searching process with stringent inclusion and exclusion criteria.The main issues explored in this study are:(1)What is the overall effect of task-based interaction on the acquisition of L2 specific grammatical features?(2)Is the effectiveness of interaction tasks related to whether the outcome measure is tested in oral or written?(3)Do different task types affect L2 acquisition differently?(4)Does the effect of task-based interaction on second language acquisition continue over time?(5)To what extent does the degree of task-essentialness impact the acquisition of specific grammatical features?The present research employed the method of meta-analysis to synthesize the findings of all experimental studies published between 2004 and 2017 to explore the empirical link of task-based interaction and L2 acquisition by coding the variables in primary studies and estimating their mean effect sizes and confidence intervals.Results of this research are as follows:(1)Overall,task-based interaction has a medium effect on the acquisition of specific L2 grammatical features which is different from a high effect gained in the previous research(Casey M.Keck et al.2003);(2)The effectiveness of task-based interaction on L2 acquisition varies with outcome measure mode.It exerts an large effect of task-based interaction on L2 acquisition in written tests while a small-level effect is found in oral tests;(3)Task type does impact the effect of interaction on second language acquisition.The type of information gap task performs the most effective,then followed by jigsaw task and narrative task exerts the smallest one;(4)The effectiveness of task-based interaction on L2 acquisition will persist over a short time;(5)Tasks in which the use of the target feature is essential exert larger effects than tasks in which the use of the target form is useful,but not required.In a word,the result of this research proves that task-based interaction can,to some extent,facilitate second language acquisition empirically,and the findings will have implications for L2 teaching practices.