Chinese High School Students' Writing Complexity, Accuracy and Fluency under GenAI Affordance in EFL Continuation Task Instruction

This study employed a repeated-measures intervention that integrated GenAI into a specific pedagogical task (continuation writing) to explore its impact on the complex system of L2 writing development in a high school context.This dataset comprises 125 longitudinal writing samples collected from 25 English learners over five consecutive writing tasks (Time 1 to Time 5). The data is organized in a long format, where each row represents a single observation, resulting in a final sample size of N=125with no missing data.The dataset includes the following six core variables:ID: A unique numerical identifier for each participant (ranging from 1 to 25), allowing for within-subject tracking across time points.Time: The measurement occasion, coded as integers from 1 to 5, indicating the sequence of the writing tasks.Syntactic Complexity: Operationalized as the Dependent Clause Ratio (DC/T), calculated as the total number of subordinate clauses divided by the total number of T-units in each composition. This variable was generated using the L2 Syntactic Complexity Analyzer (L2SCA).Lexical Complexity: Measured as the Lexical Complexity Index (LC), calculated as the sum of the most frequent 1,000 words (K1) and Academic vocabulary, divided by the total number of words. Data was extracted using VocabProfile (Lextutor).Fluency: Defined as Words per T-unit (W/T), representing the average number of words contained within each T-unit. This metric was also derived from L2SCA.Accuracy: Quantified as the Error-Free T-unit Ratio (EFT/T), calculated as the number of error-free T-units divided by the total number of T-units. This variable was obtained through manual annotation by researchers.