To Pay or Not to Pay? Investigating Students’ Willingness to Pay for ChatGPT

Abstract

Artificial intelligence (AI) is developing at a dynamic pace, playing an increasingly significant role across various sectors, including education. One widely used AI-based tool is ChatGPT, available in both a free version and a subscription-based variant offering extended functionalities. However, while the rapid adoption of such tools by students is well-documented, there remains a significant gap in the literature concerning the economic behaviours governing this new market, particularly regarding the price sensitivity of student users and the factors influencing their decision to convert from free to paid tiers. The existence of a paid version of this tool prompted a study which investigated students’ willingness to incur costs in exchange for additional features, as well as their reactions to potential price changes. The aim of this article is to determine the willingness of economics students to purchase the subscription version of ChatGPT and to analyse their responses to possible price modifications. Moreover, the study addresses the relationship between the perceived quality of the content generated by the tool and students’ willingness to pay for the premium version. A quantitative research approach was employed, based on an online survey (CAWI), in which 342 undergraduate and graduate students participated. The questionnaire included items concerning the frequency of ChatGPT usage, the evaluation of the quality of generated content, and the willingness to purchase the paid version under various pricing scenarios. Statistical methods were applied for data analysis, including measures of central tendency and correlation tests. The results indicate that ChatGPT is widely used among students – 94.4% of respondents reported using the tool, with the majority (88.9%) opting for the free version. The frequency of usage varied. Price sensitivity analysis revealed that a 25% increase in the price of the cheapest subscription would not significantly affect the decisions of most current paid users. However, a 50% price increase would result in 36.1% of them cancelling their subscription, while a 75% increase would lead to a reduction in the number of subscribers by more than half. Conversely, a 25% price decrease in the cheapest paid version would encourage 16.8% of free users to subscribe; a 50% reduction would increase this share to 42.3%, and a 75% reduction could persuade up to 70% of respondents to purchase the subscription. The article also presents the results of a Spearman’s rank correlation analysis, examining the relationship between the perceived quality of ChatGPT’s output and the willingness to purchase the paid version in the event of a price reduction. No significant evidence was found to suggest that individuals who rate the quality of ChatGPT responses more highly are more inclined to subscribe to the paid version if the price decreases.

Ähnliche Arbeiten