AI-Generated Content, Digital Misinformation, and Critical Media Literacy among Pakistani University Students: The Roles of Trust, Cognitive Reflection, and Digital Literacy
Muhammad Umar Mehmooda, Um E Rubabb, Saira Maqboolc, Muhammad Mubeen Gorayad
- aPhD Education, Punjab School Education Department
- bAssistant Professor, Department of Education, Alhamd Islamic University (AIU), Islamabad
- cVisiting Lecturer, Department of Education, Pir Mehar Ali Shah Arid Agriculture University, Rawalpindi
- dMPhil. Education, Preston University Kohat, Kohat
Papers open inside SciRah from the original publisher or repository; SciRah never hosts them.
Abstract
University students are using generative artificial intelligence to write, absorb, and evaluate information — leading to greater exposure to digital misinformation. The present research investigated the predictive value of digital literacy, critical media literacy, cognitive reflection and trust in generative AI on misinformation discernment of Pakistani university students. Methods A quantitative cross-sectional design with an embedded content-evaluation experiment. Data pooled from 1,120 students of six universities in Lahore and Rawalpindi-Islamabad using multistage stratified sampling. The participants completed measures of digital literacy, critical media literacy, AI trust and cognitive reflection, verification behaviour and a 24-item misinformation-discernment task with accurate and misleading human- and AI-generated content. Descriptive statistics, Pearson correlations, multiple regression, and repeated-measures procedures were used to analyze the data. Digital literacy had a significant positive correlation with the discernment of misinformation (r =. 419, p <. The other predictor, critical media literacy (r =. 001), was also significant. 360, p <. The second hypothesized correlation was the relationship between AI trust and discernment (H2) r = -. 343, p <. 001). Digital literacy was the strongest predictor (β =. 21), and the regression model explained 28.5% of variance in misinformation discernment, F(7, 399) = 49.76, p <.001. 256, then cognitive reflection (β =. 181) and media literacy for critical thinking (β =. 126), and AI trust served as a negative predictor (β = -. 213). Descriptively, Content which was factually correct was rated more highly than misleading content, however for misleading AI generated content, a slight preference was seen for misleading over misleading human generated content. The results underline the need for verification skills, reflective reasoning skills, and critical media literacy knowledge to develop calibrated trust in order to bolster misinformation resistance among Pakistani HEIs.
TL;DRThe results underline the need for verification skills, reflective reasoning skills, and critical media literacy knowledge to develop calibrated trust in order to bolster misinformation resistance among Pakistani HEIs. (Semantic Scholar)