Mobile-assisted ESL/EFL vocabulary learning: a systematic review and meta-analysis

Mobile-assisted language learning (MALL) has been a popular research area in recent years in the SLA field, as it has been considered as being able to facilitate language learning by offering an authentic, socially connective, contextually sensitive, and personalized mobile-mediated language learning environment. To investigate the benefits and possible limitations of mobile-assisted L2 vocabulary retention, many researchers continuously conducted relevant experiments worldwide. Among the current mobile-assisted L2 vocabulary learning research, short message services (SMS), multimedia message service (MMS), and mobile applications (apps) are the predominant approaches supported by the situated learning theory, cognitive mechanisms of L2 word retention, behaviorism, and social constructivism. Although language educators have been focusing on this topic, the effectiveness of mobile L2 vocaubulary learning still remained inconclusive. Besides, to date, little research synthesis has been conducted to investigate the effect of mobile-assisted L2 vocabulary learning. The purpose of this meta-analysis is to systematically synthethisize findings from (quaisi-) experimental studies carried out between 2005 and 2018 to examine whether there is a connection between the use of mobile technologies and L2 word retention. The results of 33 primary eligible studies showed a positive and large effect of mobile-assisted L2 word learning interventions. Additionally, SMS/MMS mode was found to be more beneficial for L2 word retention than mobile application mode; however, it is more prudent to interpret this finging as tentative because the number of target words and intervention time were unequal in the two modes. Moreover, the research settings, treatment durations, and task-afforded autonomy are three variables that tend to significantly moderate the effect of mobile-assisted L2 word retention. However, the above results should be viewed as suggestive rather than definitive with a view to the research flaws found in the primary studies and the small sample size of certain categories of moderators when compared.