A data-driven algorithm can predict which individuals are most likely to benefit from a 4-week meditation app (Healthy Minds Program) compared to no intervention. The algorithm, called a Personalized Advantage Index, was developed using machine learning on baseline data from 662 school system employees in a randomized controlled trial. It significantly moderated group differences in distress reduction, meaning it identified people who improved more with the app versus the control condition. Repetitive negative thinking alone predicted benefit nearly as well. Such an algorithm could help individuals make informed decisions about whether a meditation app is appropriate for them.
In web-based studies that track people before and after they use psychedelics, many participants stop responding, which can bias the results. Analyzing data from 654 initial participants, younger age, lower education, higher extraversion, and lower conscientiousness predicted dropping out before the four-week endpoint. Neither positive attitudes toward psychedelics nor intense challenging experiences during the drug session predicted dropout. These attrition patterns match those seen in other long-term studies, suggesting they are not unique to psychedelic research. The absence of dropout linked to psychedelic advocacy or negative drug experiences reduces concerns about certain biases in this type of data.