# Variable co-adaptation: exploring the emergence of teacher-student interaction patterns in award-winning EFL classes through the CDST lens

## TL;DR

This study explores how teacher-student interactions in EFL classes form patterns, revealing that linguistic interactions are stronger than cognitive ones, and suggests improvements for teaching practices.

## Contribution

The study introduces a CDST-based analysis of TSI in EFL contests, revealing linguistic-cognitive misalignment and proposing pedagogical improvements.

## Key findings

- Linguistic coupling in TSI was significantly stronger than cognitive coupling.
- Effective TSI elicitation correlated with higher contest rankings.
- Variability and co-adaptation drive global interaction patterns from local dynamics.

## Abstract

Teacher-student interaction (TSI) is core to effective foreign language teaching (FLT), yet its linguistic-cognitive dynamics are under-explored via Complex Dynamic Systems Theory (CDST). SFLEP Cup teaching contests form natural complex dynamic systems with fixed students and rotating teachers, offering a unique context to explore TSI emergence.

We analyzed 20-minute demo lessons of 10 award-winning EFL teachers, extracting 315 teacher-student question-answer pairs. State Space Grids (SSGs) visualized TSI dynamics; a 4-point dual coding framework was applied to video data. Spearman's correlation analysis (α = 0.05) examined bivariate associations between key TSI variables.

Individual teachers' adaptive variability converged into a dominant global TSI attractor state: simple elicitation questions paired with concise student answers. Linguistic TSI coupling (mean ρ = 0.771) was far stronger than cognitive coupling (mean ρ = 0.377). Effective TSI elicitation correlated positively with contest rankings, and variability/co-adaptation drove global pattern emergence from heterogeneous local teacher dynamics.

The low-cognitive-demand TSI pattern is shaped by contest evaluation priorities and performative teaching characteristics. Teachers designed for higher-order thinking, but student cognitive output failed to match, revealing linguistic-cognitive misalignment. Variability and co-adaptation are core mechanisms for global pattern emergence from local dynamics. We call for integrating cognitively challenging interactions into routine FLT and refining contest criteria to prioritize cognitive activation over surface-level interaction fluency, extending CDST applications in L2 classroom research.