An empirical study on effectiveness of children’s English learning ability assisted by artificial intelligence

Overview

An empirical quasi-experimental study evaluated a human-computer symbiosis teaching model designed to enhance English learning among primary school pupils. Two Grade 4 classes from a primary school in C City were randomly selected as research units; one class received the human-computer symbiosis intervention while the other continued conventional instruction. The intervention was implemented across an entire school year to assess effects on overall English achievement and multiple learning ability dimensions from an education–ecology–linguistics perspective.

Methods and approach

Two intact Grade 4 classes were assigned to experimental and control conditions through random selection of classes; pretest–posttest comparisons were conducted over one academic year. The experimental condition integrated AI-assisted tools and teacher-led activities to realize a human-computer symbiosis instructional ecology. Outcome measures included standardized English achievement scores and multi-dimensional assessments targeting language skills, comprehensive language application, higher-order cognitive ability, language strategy use, and language knowledge. Quasi-experimental analysis assessed within- and between-group changes in performance and ability dimensions.

Results

The experimental class exhibited a statistically and educationally meaningful increase in overall English achievement over the school year, while the control class showed no significant change. Performance gains in the experimental group were concentrated in language skills, comprehensive language application ability, higher-order cognitive ability, and language strategy ability. Improvement in the language knowledge dimension was minimal and did not reach practical significance. Aggregate findings indicate the human-computer symbiosis model produced differential enhancement across ability domains rather than uniform gains.

Implications

Integration of AI-enabled, human-computer symbiosis instructional designs can materially improve primary students' English achievement and higher-order and strategic language competencies within a one-year implementation. The limited effect on declarative language knowledge suggests the model may preferentially support procedural and applied language capacities; targeted curricular adjustments are required to address explicit language knowledge acquisition. Further research with larger, randomized samples, fidelity monitoring, and longitudinal follow-up is warranted to generalize findings and to disaggregate which components of the symbiotic system drive specific ability gains.