Individual differences in vision: variability rather than noise

Classically, vision scientists treat individual differences as noise and eliminate them by aggregating individuals (grand average). For example, individual-level EEG waveforms often appear highly heterogeneous. We recorded EEG during a backward masking task and found between 0 and 3 peaks in the waveforms, depending on the participant. When averaging the waveforms across participants, we obtained one peak (the classic N1), which we interpreted as the “true” waveform, while we assumed that the additional peaks “disappeared” because they were smoothed out as noise. However, when the same participants were tested again 5 and 10 years later their individual waveforms had not changed - a participant with 3 peaks still exhibited 3 peaks a decade later. Hence, individual differences in waveforms are stable traits, not noise. Importantly, many participants did not show an N1 at all. Similar results were observed in behavioural tests. Participants performed a battery of illusions. We found almost no correlations between the magnitudes of different illusions. For example, an observer may experience a strong Ebbinghaus and a weak Müller-Lyer illusion. Hence, there is no common factor for illusions. Nevertheless, the magnitude of each illusion was stable within individuals over a year. We will discuss the implications of heterogeneity for vision research in general, propose methods to measure it properly, and explore how inter-individual differences may pave new avenues in vision research.