A Meta-analysis: Shape of Age Effects and Second Language Grammar Acquisition

This study reports findings of a meta-analysis on the effects of the age of first exposure, known as the age of onset (AO), on the acquisition of morphosyntax in a second language (L2). Several studies report restrictive effects of AO on L2 grammar acquisition (Abrahamsson, 2012; DeKeyser et al., 2010; Qureshi, 2016); however, there is a disagreement about the way sensitivity to second language acquisition declines (Birdsong, 2009; Long, 2013). Researchers disagree about the nature of the decline and argue whether the decrease that occurs in learners\u27 ultimate proficiency in the second language is gradual and continuous or gradual but followed by a plateau, without further decline. To explore the nature of AO, 29 samples from 14 studies were analyzed. The results revealed a large effect size (d = 1.13, SD = .06) for the impact of AO on ultimate proficiency in L2 morphosyntax. However, for the disaggregated data (i.e., when earlier & late learners are separated), a linear decline was observed for the 6 to 16 age-group (21 samples), while a flattening of the regression line was detected for the 17 to 24 age-group (8 samples). More importantly, the findings revealed two different shapes of decline based on participants\u27 first language and the target L2. Overall, the findings support a bounded critical period for second language acquisition and indicate the typological distance between the two languages as a major influencing factor