Our perception of reality is not always perfect. Just think of how our brain reacts to optical illusions. But why is it that some people experience auditory hallucinations – say, hearing voices in their head – and others don't?
Psychiatrist A. R. Powers from Yale University and colleagues set out to find the answer. In a recently published Science study, they explain how some people are "overly influenced" by their expectations and associations they have made based on experience, and this makes them more susceptible to auditory hallucinations.
They used a technique called Pavlovian conditioning to induce auditory hallucinations. This is when one stimulus is introduced with another, creating an association between the two. In the original experiment, Pavlov trained dogs to associate the ringing of a bell with food. Here, Powers and colleagues used a checkerboard and a tone.
The researchers played the tone and display the checkerboard simultaneously to create an association between the two stimuli in the eyes (and ears) of the participants. They did this several times, adjusting the intensity of the tones. Sometimes they didn't play the tone at all and displayed the checkerboard all by itself. During the experiment, many reported auditory hallucinations (i.e. they claimed to have heard a non-existent tone) but some experienced them more often and with more intensity than others.
During the experiment, many reported auditory hallucinations (ie they claimed to have heard a non-existent tone) but some experienced them more often and more intensely than others.

Before the experiment, they had been split into four groups; two that hear voices (people diagnosed with schizophrenia and self-diagnosed psychics) and two that don’t (healthy people and people with psychosis who don’t hear voices). Each participant had to press a "yes" or "no" button, depending on if they heard the tone. The more confident they were with their decision, the longer they'd press down on the button.
Perhaps not all that surprisingly, those with a history of auditory hallucinations were more likely to "hear" a nonexistent tone. As well as being five times more likely to report a hallucination, they were 28 percent more confident in their judgment than those without a history of auditory hallucinations.
The researchers also noted that participants with a psychotic condition, whether it was schizophrenia or psychosis without auditory hallucinations, were less likely to notice when, at the end of the experiment, the tones became less frequent. This suggests that they had a harder time updating their belief about the correlation between the tone and the checkerboard compared than the others.
Computer modeling of the participants' brain turned out to be very revealing. Those who experienced severe hallucinations showed less activity in the cerebellum. The cerebellum is responsible for planning and coordinating future activity. Seeing as this requires being able to update your beliefs about the outside world, this fits with the rest of the experiment's findings.
This reaffirms previous research that has looked at the correlation between people’s assumptions and expectations with visual hallucinations. But what now? Larger studies and machine learning tactics can be used to help differentiate between psychotic and non-psychotic voice hearers. This means that, in future, researchers may be able to detect if people experiencing hallucinations are heading towards psychosis and, if they do, be able to deliver more targeted treatments.