Resumen

CHAFLOQUE-CAPUNAY, José Eugenio et al. Preliminary evaluation of the ACHRON method as a didactic model for academic writing with generative artificial intelligence. Prohominum [online]. 2026, vol.8, n.1, pp.271-288.  Epub 06-Abr-2026. ISSN 2665-0169.  https://doi.org/10.47606/acven/ph0441.

This study evaluates the ACHRON method as a critical-ethical scaffold to guide the responsible use of generative artificial intelligence (AI) in academic writing. The aim is to assess its effectiveness in improving argumentative quality and preserving authorial agency during theoretical-framework writing. A mixed-methods design with quantitative priority was implemented, using a single-group quasi-experimental pretest-posttest approach and a brief perception survey. Participants were 87 students in an online thesis-writing course (June-November 2025): 15 undergraduates and 72 graduate students (44 master’s and 28 doctoral). Academic products were assessed at both time points with a seven-dimension analytic rubric (0-28). Total scores increased from M = 13.74 (SD = 3.97) to M = 21.95 (SD = 4.03). The gain was significant, t(86) = 15.81, p < .001, with a large effect size (d_z = 1.70). Improvements were most evident in ethical AI use, originality, and alignment with the research problem. The survey showed excellent internal consistency in this pilot (α = .97; complete cases n = 85) and predominantly positive perceptions. Open-ended responses highlighted benefits for organization and traceability, alongside challenges in prompt formulation and source verification. Given the single-group design and the lack of inter-rater reliability estimation, results are interpreted as preliminary evidence consistent with ACHRON’s pedagogical value, without making strict causal claims.

Palabras clave : artificial intelligence; writing; higher education; postgraduate courses; teaching methods; educational evaluation.

        · resumen en Español     · texto en Español     · Español ( pdf )