Abstract:

This study investigates the relationship between university students’ information problem-solving (IPS) skills and their effective use of generative artificial intelligence tools, such as ChatGPT, in academic contexts. The findings demonstrate that mere familiarity with or use of AI tools is insufficient; instead, a solid mastery of IPS skills is essential for using generative AI effectively in academic tasks. Based on data from 154 Spanish university students, the results indicate that higher IPS proficiency significantly enhances students’ ability to utilize ChatGPT for solving academic information problems, while age also appears to be a relevant factor, with younger students exhibiting stronger IPS skills. The study underscores the importance of integrating IPS skills training into university curricula, as prior mastery of these skills is a key predictor of students’ competence in applying generative AI tools to academic work.

Research on AI chatbots like ChatGPT in higher education is rapidly expanding, with students increasingly favoring ChatGPT over traditional search engines for academic tasks (Bjelland et al., 2024; Zhang & Yang, 2025; Karunaratne & Adesina, 2023). While students use ChatGPT to support writing and learning, its assistance does not necessarily improve text quality, which depends on users’ prior skills (Bašić et al., 2023). The evolving nature of literacies, including information problem-solving, calls for updated assessment tools like the PIKE or PIKE-E, which remains relevant for measuring these complex skills in the AI era (Leu et al., 2017; Rosman et al., 2016; Garcia et al., 2021).

The IPS-I model is evolving to keep pace with technological changes affecting higher education, particularly the rise of AI and NLP systems like ChatGPT (Argelagós et al., 2022; Boetje et al., 2024; Sparks et al., 2024). This study achieved two main objectives: updating and validating the PIKE-E-36 test to reliably measure students’ informational competence with ChatGPT, and confirming a predictive model showing that younger students and those with stronger information problem-solving (IPS) skills use ChatGPT more effectively for academic tasks. The findings confirm that general knowledge or frequent use of AI tools alone does not ensure effective academic application; rather, pre-existing IPS skills are the key predictor (Chaudhuri & Terrones, 2025). Gender does not significantly affect ChatGPT use, but age does, with younger students displaying greater proficiency (Akter & Ahmed, 2024; Magsamen-Conrad et al., 2015). The study highlights the need for targeted IPS training in higher education, emphasizing critical thinking and ethical awareness, as these competencies are essential for leveraging AI tools effectively (Kalantzis & Cope, 2025; MacVaugh, 2025). Limitations include a small, single-country sample and reliance on self-report measures, suggesting future research should expand participant diversity and incorporate observational methods. Overall, the study underscores that strengthening IPS skills is crucial for students to fully benefit from AI in academic contexts.

Keywords:

Information problem-solving (IPS) skills, ChatGPT, higher education, academic tasks, assessment.