Evaluation of the Impact of ChatGPT on the Development of Research Skills in Higher Education

Abstract

In the context of higher education, many students struggle to develop strong research skills due to insufficient methodological training and limited academic support. Advances in artificial intelligence offer new opportunities to foster autonomous learning and improve academic writing. This study evaluates the impact of ChatGPT on the development of research skills among undergraduate nursing students in Trujillo, Peru, during the 2024 academic year. A quasi-experimental design was employed, involving 100 students divided into two groups of 50. Participants were selected through non-probabilistic convenience sampling, and group equivalence was verified through a pre-intervention diagnostic assessment. The experimental group completed a structured 15-session intervention focused on six dimensions: Problem Formulation; Information Analysis; Methodological Design; Data Analysis & Interpretation; Academic Writing; and Ethical Information Management. ChatGPT was integrated as a pedagogical support tool to facilitate idea generation, improve information retrieval, and assist with structuring academic texts. Quantitative data were collected through pretest and posttest evaluations and analyzed using Mann-Whitney U, Friedman, Bonferroni post hoc, ordinal regression, and principal component analysis. Results showed a statistically significant improvement in the experimental group (U = 95.4, p < .05), with notably increased numbers of students reaching the Very High level across all dimensions. The Bonferroni test highlighted differences among skills, and the ordinal regression model demonstrated strong predictive power (Nagelkerke R² = 0.924). These findings support the integration of ChatGPT into higher education as a strategic tool for enhancing research skills, especially in environments with limited access to traditional academic resources.

Ähnliche Arbeiten