The illusory trip: 5 Questions for psychologist and magician Jay Olson

Olson discusses magic, the power of suggestion, and placebo in psychedelic studies.

Jay Olson fell in love with magic when he was just 5 years old. By age 7, he was performing his own illusions, and as he grew up and became a psychologist, he wondered how he might use the same skills magicians do to uncover difficult-to-answer questions about what humans believe. Olson’s earliest work focused on the combination of magic and psychology: how can researchers subtly manipulate participants’ beliefs and perceptions? Now a psychology postdoctoral fellow at the University of Toronto at Mississauga, his studies have investigated how the eye reacts to illusory motion and whether clinicians can use the placebo effect to treat ADHD. In another recent study, he even built a fake brain scanning machine that could supposedly read participants’ thoughts about charity, which he used to investigate how much the sham brain scans could sway participants’ actual beliefs.

In 2020, Olson again enlisted his magic background to design a study that convinced participants they had received a psychedelic when they in fact had taken an inert pill. The Microdose spoke with him about what he found, and what it tells us about the power of suggestion and placebo in psychedelic studies.

What approach did you use to design this placebo study?

We used what I call “the Swiss cheese model of deception”: every slice of Swiss is a layer, and in every layer there are holes. If there are too few layers, these holes can line up and people can see through the deception, but magicians stack really subtle layers of deception on top of each other so no holes are exposed. They can make people believe in all sorts of things, and sometimes the deception starts well before the magic show starts.

We tried to apply that to this study: how can setting things up as a deceptive performance allow us to answer questions that we wouldn't have been able to easily answer otherwise. The placebo effect is difficult to study, and for this experiment, we wanted to show how large its effect could be if participants had the right expectations in the right kind of environment. We wanted to do everything we could to stack the deck in our favor, and to have them really believe that what they're taking is a psychedelic.

What were the layers of Swiss cheese?

We tried subtle ways of suggesting there wasn’t a placebo control group. For instance, we didn't mention anything about placebos to them; we said we were doing a “feasibility study.” We asked them to choose between pills, but all the pills were gathered from the same tray and the experimenters didn’t look at which they chose. We also used a bunch of strategies to establish credibility: there were lots of people — around 11 research assistants — involved. There was a security guard outside the door, a psychiatrist and doctor standing by, all of whom had lab coats on. We put institutional logos and labels on everything. It all set up the expectation that this was serious, and likely a psychedelic. 

So those details established expectations, and then we also carefully designed the physical and social setting to look like a stereotypical psychedelic party, which we thought would help boost any placebo effect. We also enlisted between 5 to 7 confederates to act like participants and help maintain the illusion. 

What did the confederates do?

The confederates were all friends of ours who had used psychedelics before and acted as participants. We had a training session with them beforehand and we told them to pace and lead: basically, to observe the effects that are happening in the room and then just increase them slightly. So if everybody's dancing to the music, dance a little bit harder. 

We also had them plant little statements. While the groups of participants were waiting in the lobby, there’s always a bit of nervousness and excitement, so we had one person say, “My friend did this study last week and had a blast.” 

We had somebody else who just naturally has huge pupils go up to a few people and say, “Hey, your pupils are huge — are mine like that?” We thought that maybe the participant would feel like their body was starting to have a reaction to the drug. And of course, we removed people’s smartphones so they couldn’t look at themselves using their cameras, and we even covered the mirrors in the bathroom so they couldn't check there. We were hoping that at least some of these little details would help convince some people that they had taken a psychedelic, and that expressing some of the effects might spread through the group via a sort of a contagion mechanism. 

Did it work? And how did people react when you told them it was a placebo?

People experienced a huge range of things. Some people experienced nothing whatsoever and thought it was a placebo the whole time. Some people said waves of the psychedelic experience would hit them, fade away, then come back. One person said they couldn't understand language for 20 minutes, and one person said they had a headache where gravity had a stronger hold on the back of his head; another said that he felt a tingling sensation in the chest. Some people even reported that it was indistinguishable from other psychedelic experiences they’d had in the past. 

My favorite quote from the entire study was: “So we were really sitting and staring at this painting in silence for 45 minutes, completely sober?” 

After we told them they all got a placebo, some people were shocked and surprised. A bunch of people were laughing; others were like, I knew it. In one session, a participant was accusing somebody else of being a confederate; she was an actual confederate, but we decided spontaneously during the debrief that we wouldn't tell participants who was a confederate and who wasn’t so that any participants who had been experiencing strong effects had a kind of social out. They could be like, “I was a confederate; that’s why I was acting like that.”

What does this study tell us about the potential role of placebo effects in psychedelic studies?

I think this means that some previous studies have underestimated how strong the placebo effect may be. Prior to this study, there was a lot less discussion about the placebo effect in psychedelic studies, and about control groups and unblinding. This paper may have played a small part in underlining the importance of those as challenges in study design; we showed that with the right expectations and the right context, the placebo effect can be larger than previously thought. 

We're still grappling with the best practical solutions to these challenges, though. Is it having an active placebo — people taking some other kind of drug that has some overlapping effects — versus an  inactive placebo? I feel like our paper pointed out more problems than solutions, but there have since been some innovative studies looking at really strong blinding; for instance, the study last year where they had patients under anesthesia who were either given ketamine or a placebo. Novel designs like that, where they find really large psychedelic effects and very large placebo effects, will help us understand how large the placebo effect might be in these clinical environments.

This was maybe the most fun study that I've ever run and probably will ever run. Academic research can be so boring, but I think this study shows that there's space to do crazy things that are not only fun, but also help to push the field forward in some way. 

This interview has been edited and condensed for clarity and length.