The Effect Of Computer-mediated Annotations On L2Incidental Vocabulary Acquisition:Meta-analyses

Incidental vocabulary acquisition has long been an important part in the domain of second language acquisition, and the factors affecting incidental vocabulary acquisition and what kind of factor is the most influential one has always been the top concern of many researchers. Many previous research of home and abroad has shown that as one of those factors, annotations have positive effects on incidental vocabulary acquisition, and different types of annotations have different impacts.In recent years, with the wide application of computer technology in teaching, some researchers began to draw much attention on computer-mediated annotations, namely adding other modes such as sound, pictures, video or even animation to annotations, thus further investigate the effect of such annotations on second language incidental vocabulary acquisition. As research went deep, researchers abroad began to reanalyze existing research using the method of meta-analysis. However, there is almost no researchers investigate on second language acquisition employing meta-analysis in China, let alone the meta-analysis of the effect of computer-mediated annotations on second language incidental vocabulary acquisition.Therefore, based on previous studies, this thesis explores the effect of computer-meditated annotations on second language incidental vocabulary acquisition using meta-analysis, aims to verify and expand previous researches. After the thorough search of databases of home and abroad,13studies fitting the criteria are chosen, and the meta-analyses are done. The results show that computer-mediated annotations have positive effects on second language incidental vocabulary acquisition; compared to single gloss type, multiple gloss is found to have bigger impact. When learners’proficiency level is considered, the result shows that learners at beginning level benefits the most from multiple computer-mediated annotations; as to test types, recognition tests are found to be more effective than production tests.