Seeing vs. Doing: How we measure size perception varies, but both ways show real uncertainty

What is it about?

This study looks at how consistently people perceive the size of objects, like circles, using two different measurement methods. When we study perception, it's important to know if the differences we see in people's responses are due to real differences in how they perceive things, or if they're just caused by the way we're measuring. The researchers had 180 participants do two tasks: one where they estimated the diameter of circles shown to them using a slider (a visual analog scale or VAS), and another where they were given a diameter and had to adjust a circle on the screen to match that size (a reproduction task). They used advanced statistical models to analyze how consistent people's responses were. The results showed that overall, people's responses were quite consistent across both tasks. However, there were some systematic differences. The estimation task (using the slider) tended to show greater differences between individuals and more unique responses to specific circle sizes. It also resulted in a steeper perceptual slope (meaning small changes in actual size led to bigger changes in estimated size) and a lower baseline perception compared to the reproduction task. Interestingly, while the overall amount of perceptual uncertainty (how much people's judgments varied) was similar between the two tasks, how this uncertainty changed with the size of the circle was much more noticeable in the estimation task. The findings suggest that a lot of the variability we observe in these tasks is due to genuine perceptual uncertainty, not just errors in how we measure it. Both methods are valuable for research, but they seem to tap into slightly different cognitive processes.

Featured Image

Why is it important?

This research is important because it helps us understand how reliable different methods are for measuring visual perception, specifically how people judge size. Many areas of psychology and behavioral science rely on accurately measuring perception to understand how it guides our actions and decisions, from recognizing faces to learning about threats. If the way we measure perception introduces a lot of its own errors or "noise," our conclusions about how people actually perceive the world could be wrong. This study is unique because it directly compares two common continuous measurement methods—estimation with a visual analog scale (VAS) and reproduction—and uses sophisticated modeling to break down where the variability in responses comes from. The finding that both methods show consistent patterns of variability, despite some task-specific differences, is significant. It suggests that these tasks are indeed capturing real perceptual uncertainty, which is an intrinsic part of how our brains process visual information. This work is timely as it encourages researchers to think about the cognitive demands of different perceptual tasks and paves the way for developing more precise methods to distinguish true perceptual processes from task-related noise. Ultimately, this can lead to a more accurate understanding of the link between perception and behavior.

Perspectives