Object-Based Anisotropies in the Flash-Lag Effect

Katsumi Watanabe, Kenji Yokoi
2006 Psychological Science  
The flash-lag effect is a visual effect that occurs when a stimulus is flashed in alignment with a continuously moving object resulting in the flash to appear to be lagging behind the moving object. This effect has been widely studied with the purpose of understanding why this effect occurs and to what purpose. In the review work by Bart Krekelberg and Markus Lappe (2001), both researchers analyze the different hypothetical explanations for this phenomenon. Early researchers believed that the
more » ... ash-lag effect results from the brain perceiving the flash and the motion of the object, but because of neuronal lag, has to extrapolate the position of the moving object. However, this theory is insufficient because it cannot explain the lack of predicted overshoot that would occur if the moving object were to suddenly shift position. Another explanation given is that the visual system responds with shorter latency to moving stimuli rather than to flashed stimuli. In this way, the differential latencies hypothesis states that the phenomenon is a by-product of how the visual system is setup, but the evidence is lacking that this hypothesis can explain fully the different flash-lag effects across all its different dimensions. Another school of thought, which begins to treat the actual nature of flash as special, is the temporal integration, which proposes that the visual system collects position signals over time and estimates the position. The flash is important because the information from the stimulus persists for a time and biases the position estimate towards that last scene position. Where this also diverges from the motion extrapolation hypothesis is that the temporal integration model is able to account for reversing or slowing down of the moving object and its effect on the flash lag. Finally, Krekelberg and Lappe review the Postdiction model, which explains that the brain constructs a percept by combining the 4 internal model of the world with current external input. The flash, in this way, is a high salience external input that resets the internal model of the world. After this "reset", the internal model must be rebuilt again and because the moving object has moved on, the position of the object is thought to be past the position of the flash. (Krekelberg and Lappe, 2001) Continuing in this line of research, Nijhawan (2002) explores similar explanations, but also discusses the possible biological reasons to have the primate visual system set up in this manner. There are a number of interesting effects that occur based around this flash-lag effect, which include flash-terminated and flash-initiated cycles. Flash-terminated cycle is defined as when the pre-flash trajectory of the moving object matches a regular flash lag display, but when the flash disappears so does the moving object. In this case, there is no flash lag effect. In the flash-initiated cycles, the motion of the object is started by the flash, but in this case there is a flash lag effect. In his work, Nijhawan (2002) explains that the reason for our visual system's accurate representation of position at motion-termination was evolutionarily important in how people were able to hunt and find food or resources. On the other hand, this process, also known as backward masking, of identifying flash-terminated cycles accurately seems to allow for the visual system to have some leeway in developing the flash-lag effect. (Nijhawan, 2002). In Katsumi Watanabe's and Kenji Yokoi's work (2006) , the researchers were able to find that the mislocalization of the flash was not symmetric around the moving stimulus, but rather anisotropically mislocalized. This meant that the perception of the amount of the lag of the flashing stimulus was not equal around the moving object, but rather flashed stimulus that appeared farther ahead of the moving object experienced greater flash-lag effect. Furthermore, flashed stimulus that appeared behind the moving object did not experience an equal lag effect. In fact, in their study, Watanabe and Yokoi (2006) found that the perceived flash-lag effect 5 seemed to converge onto a singular point somewhere behind the moving object. In their second experiment, Watanabe and Yokoi explored whether or not such an effect was caused by a perceived warping of the space around the object in motion. However, as the result in the experiment showed, there was no perceived warping in the space, but there was a similar effect in that it seemed the flash-lag effect was converging onto a singular point (Watanabe and Yokoi, 2006). All in all, our lab decided to focus its experiments on the flash-lag effect especially in how it deals with biological motion. As stated above, the foundation of our visual system seems to be built upon biological reasons of being able to better find resources, which may mean that biological motions, like walking or running, is processed differently than normal motion. In this study, we hypothesized that biological motion is processed differently when it comes to the flash lag effect. Method Participants There were 31 participants. They were all students from the University of California, Los Angeles that had signed up for the experiment using the SONA system. All participants received course credit for participating in the experiment. Design and Procedure All the participants were given a visual experiment task where they had to make judgments on the perceived misalignment of the flash relative to certain limbs of a walking figure that was presented on the screen. Prior to staring the experiment, the participants were asked to sit with their head about 57cm in front of the monitor they would be completing the task on. During the task, the participant would see a walking figure that was a half-green and half-6 blue and a dot would flash on the screen. The participant would either hit the left arrow key or the right arrow key dependent on which side, relative to the blue limb, the dot appeared. The dot could appear at seven different locations around the blue hand or at seven different locations around the blue leg. The dot would always appear at the tip of the walker's blue limb on frame 53 of the walker in the walking cycle. These positions were time offsets of the tip of the limbs in seconds. For example, at the 2 offset for the hand, the dot would flash at where the hand would be 2 seconds later. The experiment randomized around which limb the dot would appear, but each participant saw an equal number of dots appear either around the hand or the foot of the walker. In the experiment, there were 2 blocks of trials that the participant would go through. Each block had 140 trials for a total of 280 trials in the experiment. In the first block, the participant would either see the walker walking backwards or forward, but always the same way for the entire block. The direction of the walker would flip going into the second block. Participants were randomly placed into which condition (forward or backward walking walker) they saw first. Before the actual experiment, there were 10 practice trials that had trials exactly like the ones in the testing phase, but these trials ran about two times slower during the practice trial. At the very end, all participants were given a questionnaire to find the participant's autistic quotient score. This was key because there has been previous work done suggesting that autism can affect the ways that people perceive biological motions (van Boxtel and Lu, 2013). Results and Discussion As seen in Figure 1 (below), each line is shifted toward the positive time offset, which is evidence of the flash lag effect because it demonstrates that the participants on average were 7 saying that the flashed dot appeared on the opposite side of the limb than it actually did appear. The results demonstrate that participants demonstrated a larger flash-lag effect for the foot going forward than when the foot was going backward. On the other hand, there doesn't seem to be the same significant difference in effect when it comes to the hands whether going forward or backward. Such results may suggest that there are special mechanisms within the visual system that deal with the motion of legs, but why it has such a larger effect on forward moving feet is unknown. With these results, future studies will focus on explaining why there is such a large effect on the forward moving feet rather than when the backward moving feet and why this is not seen in the hand movements. Furthermore, future studies will include studying the effects of speed on the flash lag effect as limbs do not always move at a constant speed naturally. Figure 1. Plots of average proportion of responses given that corresponded to the direction of the motion of the walker. One can see that there is a significant increase in responses for the 8 forward-facing walker's foot at all positive time offsets as compared to the responses for the backward-facing walker's foot. 9 References Krekelberg, B. and Lappe, M. (2001). Neuronal latencies and the position of moving objects. TRENDS in Neurosciences, 24(6). 335-339. Nijhawan Romi (2002). Neural delays, visual motion and the flash-lag effect. TRENDS in Cognitive Sciences, 6(9). 387-393. van Boxtel, J. J. A and Lu, H. (2013). Impaired global, and compensatory local, biological motion processing in people with high levels of autistic traits. Frontiers in Psychology, 4.
doi:10.1111/j.1467-9280.2006.01773.x pmid:16913957 fatcat:v5vmnxzecbdypjdpq57pezvmze