Exploring the potential of Generative AI as a support tool for single-case data analysis

Visual analysis is the primary method used to analyze data in single-case experimental designs (SCEDs). Yet, whether valid conclusions can be drawn from visual analysis remains a concern. There is a need to explore new methods that not only are able to make accurate conclusions from SCED data graphs but are also easy to use. In the present study, we extended the findings of Akuoko-Mensah and Chen (2025) and examined the capabilities of ChatGPT and Claude for analyzing SCED graphs. SCED graphs available at www.singlecase.org were used to train and test ChatGPT and Claude. The findings revealed that Claude's ratings closely aligned with expert ratings on both ABAB and multiple-baseline graphs. In contrast, ChatGPT's ratings were closely aligned with expert ratings on ABAB graphs only. In addition, the graph descriptions and rating justifications generated by Claude included more quantitative detail and terminology related to specific design forms. Based on these findings, we discuss study implications and future research directions.