Influence of Linguistic Regularities on Instructional Behavior and Pedagogical Evaluation

Ascertain whether linguistic regularities documented in large language model–generated text shape instructional behavior and pedagogical evaluation in tutoring contexts.

Background

Prior work has identified consistent linguistic patterns in model-generated language across domains, but these analyses typically do not tie language differences to task-specific evaluation criteria or instructional outcomes.

The authors explicitly note uncertainty about whether such linguistic regularities influence instructional behavior or how feedback is evaluated within educational tutoring settings, motivating their analysis of instructional and linguistic features alongside pedagogical quality.

References

As a result, it remains unclear whether such linguistic regularities shape instructional behavior or pedagogical evaluation in tutoring contexts.

— Large Language Models Approach Expert Pedagogical Quality in Math Tutoring but Differ in Instructional and Linguistic Profiles  (2512.20780 - Abdulsalam et al., 23 Dec 2025) in Section 1, Introduction

Also, even though we consider implicit conditioning, the fact that we experiment on English only implies that we cannot properly evaluate the effect of grammatical gender, which is much more significant in other languages .

— The Role of Implicit and Explicit Demographic Signals in Large Language Model-based Student Assessment  (2609.16993 - Rooein et al., 15 Sep 2026) in Limitations section