Psychological researchers have used AI to not only test a hypothesis on COVID-related behaviors but also to help come up with the hypothesis in the first place.
The theorem was that individuals who feel optimistic about the future of humanity will behave more ethically—not hoarding PPE or flouting social distancing—than their pessimistic peers.
The research was conducted by scientists in Singapore and China and published in Psychological Science.
The team trained a deep learning model to tease out the hypothesis, then tested it on two cohorts of U.S. residents whose combined number topped 500.
“The deep neural network that we built was able to classify respondents of the World Values Survey as ethical or unethical with high accuracy—the model’s accuracy was 90% of the test-retest accuracy of the measure of unethicality,” report Abhishek Sheetal of Nanyang Technological University and colleagues.
Commenting on the work for The New York Times, psychologist Wiebke Bleidorn, PhD, of UC-Davis says the study “highlights that a theory-blind, data-driven search of predictors can help generate novel hypotheses. And that theory can then be tested and refined.”
The study is available in full for free.