Data set supplementing "Technology-supported Self-triage Decision Making"

This is the de-identified data set used to conduct the analyses of our study "Technology-supported Self-triage Decision Making" (https://doi.org/10.1038/s44401-024-00008-x). 

This dataset contains 600 respondents' appraisals of real case vignettes (built using the RepVig framework). After their initial appraisal, participants received advice from either the Large Language Model (LLM) ChatGPT or from the Symptom-Assessment Application (SAA) Ada Health and made a self-triage decision again. They also reported their corresponding certainty for both decisions.  

Additionally, the data contains participants'

education

medical/first aid training

previous experience with SAAs and LLMs

Minimum European Health Module (general health, chronic morbidity, activity limitations)

self-efficacy (measured using the Allgemeine Selbstwirksamkeit Kurzskala)

technology affinity (measured using the Affinity for Technology Interaction Short scale. Note: items 3 and 4 are reverse coded)