TY - JOUR
T1 - A meta-analysis of syntactic priming experiments in children
AU - Kumarage, Shanthi
AU - Donnelly, Seamus
AU - Kidd, Evan
N1 - Publisher Copyright:
© 2024 The Author(s)
PY - 2024/10
Y1 - 2024/10
N2 - A substantial literature exists using the syntactic priming methodology with children to test hypotheses regarding the acquisition of syntax, under the assumption that priming effects reveal both the presence of syntactic knowledge and the underlying nature of learning mechanisms supporting the acquisition of grammar. Here we present the first meta-analysis of syntactic priming studies in children. We identified 37 eligible studies and extracted 108 effect sizes corresponding to 76 samples of 2,378 unique participants. Our analysis confirmed a medium-to-large syntactic priming effect. The overall estimate of the priming effect was a log odds ratio of 1.44 (Cohen’s d = 0.80). This is equivalent to a structure that occurs 50 % of the time when unprimed occurring 81 % of the time when primed. Several variables moderated the magnitude of priming in children, including (i) within- or between-subjects design, (ii) lexical overlap, (iii) structural alternation investigated and, (iv) the animacy configuration of syntactic arguments. There was little evidence of publication bias in the size of the main priming effect, however, power analyses showed that, while studies typically have enough power to identify the basic priming effect, they are typically underpowered when their focus is on moderators of priming. The results provide a foundation for future research, suggesting several avenues of enquiry.
AB - A substantial literature exists using the syntactic priming methodology with children to test hypotheses regarding the acquisition of syntax, under the assumption that priming effects reveal both the presence of syntactic knowledge and the underlying nature of learning mechanisms supporting the acquisition of grammar. Here we present the first meta-analysis of syntactic priming studies in children. We identified 37 eligible studies and extracted 108 effect sizes corresponding to 76 samples of 2,378 unique participants. Our analysis confirmed a medium-to-large syntactic priming effect. The overall estimate of the priming effect was a log odds ratio of 1.44 (Cohen’s d = 0.80). This is equivalent to a structure that occurs 50 % of the time when unprimed occurring 81 % of the time when primed. Several variables moderated the magnitude of priming in children, including (i) within- or between-subjects design, (ii) lexical overlap, (iii) structural alternation investigated and, (iv) the animacy configuration of syntactic arguments. There was little evidence of publication bias in the size of the main priming effect, however, power analyses showed that, while studies typically have enough power to identify the basic priming effect, they are typically underpowered when their focus is on moderators of priming. The results provide a foundation for future research, suggesting several avenues of enquiry.
KW - Language acquisition
KW - Meta -analysis
KW - Syntactic priming
KW - Syntax acquisition
UR - http://www.scopus.com/inward/record.url?scp=85193290257&partnerID=8YFLogxK
U2 - 10.1016/j.jml.2024.104532
DO - 10.1016/j.jml.2024.104532
M3 - Article
SN - 0749-596X
VL - 138
SP - 1
EP - 31
JO - Journal of Memory and Language
JF - Journal of Memory and Language
IS - October
M1 - 104532
ER -