A new study has revealed large differences in attitudes towards various types of drugs, with a growing acceptance of psychedelics and strong negative feelings towards more traditional illicit drugs.
Psychology experts from the University of Nottingham analysed over 600,000 Reddit posts using Google Cloud Natural Language Infrastructure to compare how different substances were talked about online in terms of positivity versus negativity.
The study, published in the Journal of Psychoactive Drugs, provides new insights that could help clinicians and researchers shape educational materials and design trials that address real concerns around different types of drugs.
Understanding public attitudes toward emerging treatments is crucial, as these perceptions shape research priorities, inform clinical practices, and guide regulatory decisions. Social media platforms, such as Reddit, offer a rich, first-person window into how people describe their experiences, share opinions, and debate potential benefits or risks, insights that traditional surveys often miss. Because Reddit users remain largely anonymous, many feel freer to share candid, first-person accounts of their experiences.
The researchers observed the distinct sentiment profiles across substances: traditional illicit drugs like heroin and methamphetamine consistently evoked strongly negative responses, underscoring enduring public stigma and concern.
Psychedelics, such as psilocybin (found in "magic mushrooms") and LSD, are psychoactive compounds that, when combined with guided therapy, often called psychedelic-assisted therapy, have shown promise in treating conditions like depression, anxiety, and PTSD. These types of drugs tended to generate neutral-to-slightly positive sentiments, perhaps reflecting a growing acceptance of their potential therapeutic value, especially notable for ayahuasca and mescaline, which drew the most positive emotional responses among all compounds studied.
PhD student Brandon Biba is the lead author on the study, he adds: "Natural language processing enables a scalable, objective analysis of vast amounts of unstructured text, automating the identification of key themes and overall sentiment in a fraction of the time required for manual coding."
Our study demonstrates that tracking public sentiment around psychedelics gives the field invaluable guidance. Relying solely on surveys or interviews to learn about people's experiences and opinions introduces social desirability and recall bias, where people often give the answers they think we want to hear, or simply forget details. By contrast, scanning social media lets us tap into casual, informal discussions that may reflect more honest viewpoints. Yet, neither approach tells the whole story on its own; we need both formal surveys and informal online data to really understand where people stand.
Google Cloud's NLP tools allowed the researchers to sift through massive amounts of text in hours, automatically pulling out themes, measuring sentiment intensity, and showing exactly how each substance is discussed. This kind of automated analysis cuts down on time, cost, and human bias, and it scales easily, and could potentially expand to other discussion platforms and build a more comprehensive view of attitudes worldwide.
Dr O'Shea adds: "By combining large-scale data with the detailed findings from our study, everyone, from researchers to regulators, can help inform decisions to ensure new therapies are aligned with what communities really need."
