The aim of this study is to comprehensively evaluate quantitative experimental mobile-assisted language learning (MALL) studies published between 1994 and 2019 that meet minimal conditions of research design and statistical analysis. Starting with a bibliographical database of 1,144 references to experimental MALL implementations, of which there were 700 objectively substantiated by quantitative experimental language learning outcomes, only 84 experimental MALL studies met the inclusion requirements. Their analysis addresses two critical sets of research questions. First, what are the general characteristics of the selected studies and, second, what are their language learning outcomes in terms of measured effect size. Nine general characteristics are considered: publication source, chronological distribution, country of origin, institutional environment, sample size, intervention duration, targeted language, language learner competence level, and learning focus. Effect size was calculated separately for between-group (independent, experimental) and within-group (quasi-experimental) treatment studies. In both cases, the overall results were quite large: 0.72 for the former and 1.16 for the latter. An analysis of four critical moderator variables (language learner competence level, language area focus, institutional environment, and intervention duration) revealed similarly large effect sizes. Notwithstanding, analysis of the data also confirmed obvious publication bias and a very high level of heterogeneity that frequently approached 100%. The relevance of positive language learning outcome conclusions thus needs to be tempered by these shortcomings.

In order to build complex language from perceptual input, children must have access to a powerful information processing system that can analyze, store, and use regularities in the signal to which the child is exposed. In this article, we propose that one of the most important parts of this underlying machinery is the linked set of cognitive and language processing components that comprise the child's developing working memory (WM). To examine this hypothesis, we explore how variations in the timing, quality, and quantity of language input during the earliest stages of development are related to variations in WM, especially phonological WM (PWM), and in turn language learning outcomes. In order to tease apart the relationships between early language experience, WM, and language development, we review research findings from studies of groups of language learners who clearly differ with respect to these aspects of input. Specifically, we consider the development of PWM in children with delayed exposure to language, that is, children born profoundly deaf and exposed to oral language following cochlear implantation and internationally adopted children who have delayed exposed to the adoption language; children who experience impoverished language input, that is, children who experience early bouts of otitis media and signing deaf children born to nonsigning hearing parents; and children with enriched early language input, that is, simultaneous bilinguals and second language learners.