Limitations
Although a strength of the study is the large sample, we relied on an online format. This left uncontrolled a host of viewing parameters (e.g., image resolution, lighting, person-to-screen distances, screen dimensions) we would prefer to control. These sources of noise are likely equivalent across participants, and we see significant effects. The added noise may weaken effects compared to in person testing. This is worth revisiting with in person testing going forward. Online data collection prevented us from linking visual illusion findings in SSD with neural mechanisms (e.g., Thakkar et al., 2019). The number of illusions seen in the PGT has been associated with cortical hyperexcitability in visual cortex (Braithwaite et al., 2015; Fong et al., 2020), though we did not measure this in the current study, and cannot confirm this in our samples. It is possible that cortical hyperexcitability/hyporeactivity fluctuates across the cortex.To capture the underlying mechanism accounting for the relationship between high schizotypy and illusory percepts we are now collecting neural responses during task performance. The absence of neural measures leaves open an alternative hypothesis that there is a signal detection difference that shifts response criterion in those with higher schizotypy scores. In other words, those with high schizotypy might bemore inclined to report that an illusion was perceived than those with lower scores (Sahakyan & Kwapil, 2019), and thus we could be capturing a response bias rather than a difference in perceptual experiences.