Exploring Student Preference between AI-Powered ChatGPT and Human-Curated Stack Overflow in Resolving Programming Problems and Queries

Abstract

Among computer programmers and developers, the user-oriented question-and-answer website of Stack Overflow is a useful platform for sourcing solutions to programming problems, exchanging insights, and accessing a wealth of shared knowledge. However, the timeliness of responses on this platform is frequently a limiting factor that ChatGPT could potentially address. The goal of this study was to explore the preferences of novice programmers between these platforms for finding answers to their programming questions. Anchored in the Technology Acceptance Model (TAM) and the Information Foraging Theory (IFT), the study investigates users' perceptions of usefulness, ease of use, information scent, cognitive effort, as well as overall preferences. Our findings show discernible variations in preferences within the group of students (i.e., application and website developers). In line with these results, we discussed theoretical and practical implications and suggested a dual-pronged approach to leverage both environments as coding assistants in computer programming education. © 2024 Elsevier B.V., All rights reserved.