“ChatGPT Seems Too Good to be True”: College Students’ Use and Perceptions of Generative AI

Abstract

This survey study investigates ChatGPT’s usage and perceptions among U.S. college students (N=1001), exploring its relationship with societal structures and values. Most participants reported using ChatGPT at least monthly for general tasks and writing, while its use for computer programming was minimal. Non-native English speakers utilized ChatGPT more frequently for writing, consistent with its perceived role as a leveling tool in academic language challenges. Natural Language Processing and thematic analysis revealed five main student concerns about using ChatGPT: ethical implications, skepticism towards AI, quality of ChatGPT’s output, potential loss of learning and job opportunities, and accessibility. While some students expressed no concerns, citing confidence in their appropriate use of ChatGPT, others hesitated to use it for writing, valuing their unique writing process and voice. Computer science majors expressed concerns related to job displacement due to the advent of generative AI. Students from low-income backgrounds highlighted constraints in resources as a barrier to effectively utilizing ChatGPT. Sentiment analysis revealed that students from higher-income groups generally held less negative views on ChatGPT compared to their lower-income counterparts. Our research underscores how technology can both empower and marginalize within educational settings; we advocate for equitable integration of AI in academic environments.

Ähnliche Arbeiten