Reanalyzing L2 Preposition Learning with Bayesian Mixed Effects and a Pretrained Language Model

arXiv cs.AI / 4/6/2026

💬 OpinionIdeas & Deep AnalysisModels & Research

共有:

Key Points

The study reanalyzes Chinese learners’ English preposition performance data using both Bayesian mixed-effects models and neural modeling approaches.
It largely replicates earlier frequentist results while also uncovering new interactions involving learners’ ability, task type, and the specific stimulus sentence.
The authors argue Bayesian methods are especially valuable given the dataset’s sparsity and the high diversity of learners.
The work suggests a promising direction for using pretrained language model probabilities as predictors of grammaticality and learnability.

Abstract

We use both Bayesian and neural models to dissect a data set of Chinese learners' pre- and post-interventional responses to two tests measuring their understanding of English prepositions. The results mostly replicate previous findings from frequentist analyses and newly reveal crucial interactions between student ability, task type, and stimulus sentence. Given the sparsity of the data as well as high diversity among learners, the Bayesian method proves most useful; but we also see potential in using language model probabilities as predictors of grammaticality and learnability.