Modeling the determinants of English writing performance: Directions to interdisciplinary writing instruction

Abstract

This cross-sectional study examines the impact of five learner-factor groups – intellectual, learning, social, environmental, and mental domains, on writing performance using confirmatory factor analysis and multiplegroup structural equation modeling, and qualitative analysis of response scripts. Triangulated data was collected from the persuasive-writing task scores and the Likert-scale and interview-based questionnaire responses of 499 first-year Thai undergraduates from 11 faculties at a university in central Thailand. The results showed the global fits between the hypothesized model and the empirical data (Chi-Square = 330, df = 169, p-value < .001, CFI = 0.96, TLI = 0.95, RMSEA = 0.04). The intellectual appeared to be the most powerful factor affecting participant writing performance. The learning factors which impacted the writing performance of the Science and Technology group (β = -.24, t = -2.60) with the highest degree of significance was out-of-class activity (β = .81, t = 42.12, p ≤ .05). Intellectual factors were shown to affect female student writing performance (β = .62, t = 9.42) more so than males (β = .44, t = 6.14, p ≤ .05). Participants viewed instrumental motivation as affecting their writing performance (mean = 2.98, SD = .93, p = .04). Qualitative data from gender non-conforming participants’ responses also uncovered underlying factors: remote student-teacher relationships and unwillingness to communicate, impairing their English learning and writing performance. Insights gleaned from the responses of gender non-conforming participants have led to suggestions for further research regarding writing instruction.