Using Visual Illusions to Determine How Individuals Perceive the Present


Motion-induced position shifts (MIPS) are a group of visual motion illusions where an object’s position is biased by motion, causing the object’s position to be misperceived. A notable example of a MIPS is the flash-lag effect (Nijhawan, 1994). In the flash-lag effect, a stationary object briefly presented in physical alignment with a moving object, is perceived in a position behind the moving object, even though the stationary object and moving object are aligned. Previous research has proposed several mechanisms (e.g., prediction and postdiction) of the brain that may underly MIPS (Eagleman et al., 2007; Hogendoorn et al., 2020). Unfortunately, due to past literature’s tendency to individually study these illusions, there is minimal agreement on which mechanisms are responsible for this group of illusions. Currently, the literature seems to suggest that most illusions are caused by different mechanisms. As these illusions all have similar illusory effects, intuitively it would be expected that some of these illusions are due to shared mechanisms. By studying if MIPS are due to shared or discrete mechanisms, we can improve understanding of the mechanisms underlying motion perception. Visual perception and illusory perception are not consistent across individuals, everyone has their own unique percept. Using an individual differences approach, we can explore the extent to which shared mechanisms underly MIPS, as individuals should have a similar strength of illusory effect to illusions that are due to shared mechanisms (Grzeczkowski et al., 2017). Please see this link for a demonstration of the flash-lag effect:

Research Questions / Hypotheses

Past MIPS literature using an individual differences approach only used two illusions (Morrow & Samaha, 2021). By only studying two illusions, it contributes limited knowledge to our understanding of the mechanisms underlying all MIPS. Addressing this gap, the present study uses an individual differences approach with eight illusions to address the research question: “are motion-induced position shifts due to shared or discrete mechanisms?”. Using correlation analyses to address this question, we would expect that illusions with positively correlated illusory magnitudes likely share mechanisms. Whereas illusions that do not correlate likely have discrete mechanisms.


Data collection and analysis is ongoing. In semester 1, 45 REP participants participated in this study. Of these, only 41 participants completed two sessions. All participants were right-handed, had correct or corrected to normal vision, and no history of a neurological disorder. Currently, participants were excluded from an experiment’s analysis if they pressed the same key for 80% of a condition’s trials, if their results indicate they did not understand the experimental task, or if they failed more than 20% of the experiment’s attention checks. Participants missing data for more than half of the experiments will be excluded from analysis.


Participants completed two two-hour sessions. In each session, in an enclosed dark room, participants rested their head on a chinrest while they viewed eight MIPS displayed on a computer monitor. Participants viewed: the flash-lag effect (luminance and motion), flash-drag effect, flash-grab effect, flash-jump effect, stationary objects containing internal motion, Fröhlich effect, and the twinkle-goes effect. Before each experiment, participants completed a short Qualtrics survey on the experiment’s instructions. During all tasks except the luminance flash-lag effect and the flash-jump effect, participants used a keyboard or mouse to indicate the perceived position of an object. During the luminance flash-lag effect, a static circle was flashed alongside a circle changing in luminance (becoming brighter or darker); participants reported which circle was darker. In the flash-jump effect, participants used a keyboard’s arrow keys to adjust the height of one of two horizontally moving black bars changing in height (either growing or shrinking), to ensure these bars were the same height when they flashed white.


Data collection, cleaning, and analysis are ongoing. Preliminary descriptive statistics indicate that for all illusions, illusory magnitude slightly decreased from session 1 to 2. A correlation analysis found there were large positive correlations between the illusory magnitudes of session 1 and 2, suggesting illusory magnitude was stable across time. In the future, t-tests will be conducted exploring whether this decrease in magnitude is statistically significant. Because illusory magnitude appears stable across sessions, illusory magnitude was averaged across both sessions. Spearman’s Rho correlation analyses on average illusory magnitude only identified 4 significant positive correlations, 3 of these correlations involved three illusions, which were all correlated with one another. Data collection is ongoing and these results are likely to change. We plan to correct the correlation analyses for multiple comparisons.


The correlation analyses indicated that three illusions are positively correlated with one another. This suggests that these three illusions may be due to the same shared mechanisms. The absence of correlations between many illusions, suggests that many MIPS illusions may be due to discrete mechanisms. Ultimately, these results imply that MIPS and motion perception may be caused by shared and discrete mechanisms. The preliminary findings of this study are being presented as a poster at the 2022 Australasian Brain and Psychological Sciences Meeting. We intend to communicate the final results in journal articles and a PhD thesis.