Assessing novice programmers' perception of ChatGPT:performance, risk, decision-making, and intentions

TL;DR

This study investigates how novice programmers perceive ChatGPT's performance, risks, and decision-making influence, revealing that positive perceptions and risk-reward assessments significantly boost their intention to adopt ChatGPT for programming tasks.

Contribution

It provides empirical evidence on the factors influencing novice programmers' adoption of ChatGPT, emphasizing the roles of performance expectancy, risk-reward appraisal, and decision-making.

Findings

Higher performance expectancy improves decision-making.

Favorable risk-reward perception increases confidence in using ChatGPT.

Positive perception of decision-making role boosts usage intention.

Abstract

This study explores the novice programmers' intention to use chat generative pretrained transformer (ChatGPT) for programming tasks with emphasis on performance expectancy (PE), risk-reward appraisal (RRA), and decision-making (DM). Utilizing partial least squares structural equation modeling (PLS-SEM) and a sample of 413 novice programmers, the analysis demonstrates that higher PE of ChatGPT is positively correlated with improved DM in programming tasks. Novice programmers view ChatGPT as a tool that enhances their learning and skill development. Additionally, novice programmers that have a favorable RRA of ChatGPT tend to make more confident and effective decisions, acknowledging potential risks but recognizing that benefits such as quick problem-solving and learning new techniques outweigh these risks. Moreover, a positive perception of ChatGPT's role in DM significantly increases the inclination to use the tool for programming tasks. These results highlight the critical roles of perceived capabilities, risk assessment, and positive DM experiences in promoting the adoption of artificial intelligence (AI) tools in programming education.