Data set supplementing "Increasing Large Language Model Accuracy for Care-Seeking Advice Using Prompts Reflecting Human Reasoning Strategies in the Real World: Validation Study"

This is the de-identified data set used to conduct the analyses of our study "Increasing Large Language Model Accuracy for Care-Seeking Advice Using Prompts Reflecting Human Reasoning Strategies in the Real World: Validation Study" (https://doi.org/10.2196/88053)

This dataset contains data collected from ten OpenAI large language models on care-seeking advice. Each model was prompted with 45 case vignettes, 10 times per vignette and model, and includes three prompts: a default prompt, a prompt inspired by the data-frame theory, and a prompt inspired by recognition-primed decision-making. Next to the model response, it also includes the correct solution of each case.