We conducted a systematic review and meta-analysis on the association between nonword repetition (NWR) and language exposure in bilingual children and explored whether the association is influenced by other variables. We performed a blind literature review on ERIC and Google Scholar, a random-effects model meta-analysis and subgroup analyses to test potential moderators. Out of 822 screened articles, we identified 24 works including 1399 children. Significant associations were found using either cumulative or current exposure, language-like nonwords, phoneme NWR scoring, in children with typical language development. Nonsignificant associations were found in studies either using age of first exposure, on children older than six, with less than 50 participants, using NWR lists containing 16-24 nonwords or with participants having different native languages. Weak associations were found when considering whole-word scoring or gray literature. We highlight the contributions of different variables to NWR, and evidence to optimally design NWR for bilingual language assessment.

我们对双语儿童的非单词重复 (NWR) 与语言接触之间的关联进行了系统回顾和荟萃分析，并探讨了该关联是否受其他变量的影响。我们对 ERIC 和谷歌学术搜索进行了盲法文献回顾，随机效应模型荟萃分析和亚组分析来测试潜在的调节因素。在筛选的 822 篇文章中，我们确定了 24 篇作品，其中包括 1399 名儿童。在具有典型语言发展的儿童中，使用累积或当前接触​​、类似语言的非词、音素 NWR 评分发现了显着关联。在使用首次接触年龄、6 岁以上儿童、参与者少于 50 人的研究中发现了不显着的关联，使用包含 16-24 个非单词的 NWR 列表或与具有不同母语的参与者一起使用。在考虑全词评分或灰色文献时发现了弱关联。我们强调了不同变量对 NWR 的贡献，以及为双语语言评估优化设计 NWR 的证据。