Measuring student receptivity to ChatGPT in higher education: A case study from Peru

Abstract

The continuous rise of artificial intelligence tools in educational settings highlights the importance of understanding how students position themselves in relation to these technologies. This study aims to analyze the attitudes toward ChatGPT among Peruvian university students, considering the cognitive, affective, and behavioral components, as well as the metric properties of the instrument. A quantitative approach was employed, with a descriptive-comparative and instrumental research design, involving 464 Peruvian students from different academic disciplines. The results indicated that, after performing confirmatory factor analysis, the final model consisted of 33 items assessing attitudes toward ChatGPT across cognitive, affective, and behavioral components, with slightly acceptable fit indices (χ²/df, p < 0.05, SRMR, RMSEA, TLI, CFI, and GFI) and adequate factor loadings (λ > 0.3). In addition, the instrument showed satisfactory reliability evidence (α<sub>ordinal</sub> and Ω<sub>ordinal</sub> > 0.7). Another finding revealed that engineering students exhibited a significantly more favorable affective attitude toward ChatGPT (p = 0.03) compared to students from the social and natural sciences. No significant differences (p > 0.05) were found in students’ attitudes based on sex or age. In conclusion, although future engineers display a more favorable affective attitude than students from other disciplines, overall attitudes toward ChatGPT do not show relevant differences across other sociodemographic factors. Moreover, the instrument proved to be valid and reliable with 33 items, thus representing a solid and less dense tool for future research.

Ähnliche Arbeiten