Using AI to Translate Medical Outcomes and their Probabilities for the Benefit of the Public

Facilitated by AI, this project builds young people’s capacity to examine data about medical outcomes for themselves. These skills will be designed to complement learning from formal education (e.g., high school curricula). Thus, the target population will be youth to enhance lifelong skills in critical thinking that are grounded in validated scientific research. The project will familiarize students with cutting-edge uses of artificial intelligence (AI) models to access and explain technical scientific information to the lay public.

The CCE intern will develop skills in AI, such as prompt engineering, that promote understanding of scientific standards for clinical trials’ outcomes (efficacy) and probabilities of adverse events (safety). We will test the effectiveness of this translation process by assessing (a) understanding of cause and effect in medical trials, (b) interpreting magnitudes of probabilities (using varying formats ranging from categorical to numerical), and (c) changes in what is called "gist plausibility"; gist plausibility is based on a theoretically motivated and validated scale to measure positive transfer to beliefs about the plausibility of health claims. In contrast to true/false or agree/disagree judgments, the plausibility approach should enhance young people’s ability to generalize from the specific facts they have been presented with to other domains of health.

Roles and responsibilities

Our intern will help us develop and run the study among adolescents participating in CCE programming, including 4-H, and Cornell Undergraduates. The intern will work closely with Dr. Reyna’s research team and meet with our CCE partners to extract content from scientific articles about medical outcomes, critically, adverse outcomes. Responsibilities will include, but are not limited to, meeting with the project team (including CCE partners), stimuli development and validation, survey design and delivery methodology, data entry, and data analysis.

Qualifications and previous coursework

This opportunity is available to students in Cornell University's College of Human Ecology.

The preferred candidate will demonstrate a high level of enthusiasm for Dr. Reyna’s research in risk communication and medical decision making, as well as for applying this research in outreach and educational settings, including working with youth. The candidate will also have completed general coursework in at least one of the following: Human Development; Psychology; Human Biology, Health and Society; Neurobiology; or related fields. The candidate should be in excellent academic standing.

Learning outcomes

The intern will have the opportunity to gain skills in program development, translation and application of research to real-world problems, working with others, translating research into outreach and educational materials, and youth development programming. They will also have the opportunity to gain knowledge of psychology and related behavioral sciences, medicine, public health, artificial intelligence (e.g., large language models and implementations of generative artificial intelligence models), and issues in education.