The Effects of Autism on Change Blindness

Although people generally have a good memory for visual information, they sometimes fail to see dramatic changes (Simons, 1996). Change Blindness is a phenomenon that occurs when changes in a scene are hard to perceive or are not perceived at all. Even when the observer knows in advance to look for changes, and when the changes are seemingly obvious and occur frequently, they are literally blind to these changes that are taking place. (Rensink, O’Regan, & Clark, 2000). While there is no set theory that fully explains change blindness, much research has been done in an attempt to understand how and why it happens.

There are several methods of testing the phenomenon of change blindness, including flicker and video paradigms. In a video paradigm, an element of the scene changes as the camera angle changes. For example, in one scene a woman was wearing a pearl necklace, and in the scene immediately after it she was wearing a scarf instead. In a flicker paradigm, an original image is repeatedly switched back and forth with a modified version of the same image, while blank fields are positioned between subsequent images. The participant views the flickering image until they see the changes (Rensink, O’Regan, & Clark, 2000).

One such experiment involved the use of a flicker paradigm in an attempt to find the cause of change blindness. Rensink, O’Regan, and Clark performed a series of experiments that manipulated the length of viewing time (Experiment 1), blank durations (Experiment 2), blank colors (Experiment 3), and the amount of interruption (Experiment 4).

It was hypothesized that change blindness might be due to volatility, which refers to the instability of a scene. If a scene is volatile, our visual system cannot perceive the changes in the scene because focused attention is required to do so. Without focused attention, coherent visual representations cannot be formed. A second hypothesis was that change blindness might be due to disruption, in which case interrupting the scene with blank cards is what caused the change blindness to occur in the first place.

The results of Experiment 1 showed that extra time to view the image does not reduce change blindness, because participants in the extended preview group scored about the same as the control group. This implies change blindness is not caused by a lack of time to construct a meaningful representation. In Experiment 2, change detection was relatively low for both the control group and the experimental group, although response times were faster for the group that were shown the blank for the least amount of time. In Experiment 3, three different color blanks were used: Black, white, and red. (The standard blank color is gray). A greater degree of change blindness was induced by the red blank fields, although the cause for this is not entirely known. Experiment 4 showed that responses were slower for the non-interuption condition, which proves that focused attention is needed to detect changes. (Rensink, O’Regan, & Clark, 2000). While some researchers attribute change blindness to volatility or interruption, other researchers have entirely different hypotheses as to what causes change blindness.

It was postulated by Simons that change blindness occurs because object representations are not sustained across views. In other words, our perceptual system takes for granted that characteristics of objects or scenes stay the same. Since we assume that what is seen is the same as what we just saw, we don’t expect to see any changes. In this memory task experiment, participants were required to watch arrays and state whether they were the same or different. The effect of labeling versus not labeling the object was also measured, in order to determine how encoding influences change blindness. The results supported the hypothesis, thus, encoding is necessary to notice changes. (Simons, 1996). While there are many theories as to why change blindness occurs, there is not one that fully explains the occurrence of change blindness.

Much research has been done on the visual perceptions and disturbances that accompany autism. Understanding impaired perception is important for facilitating an understanding of “normal” perception (Sekuler & Blake, 2002). Autism is a disorder characterized by social withdrawal (Kanner, 1943), as well as “non-social perceptual and attentional deficits” (O’Riordan, Plaisted, Diver, & Baron-Cohen, 2001). Individuals with autism also have a striking ability to notice features and changes in the environment that may go unnoticed by people that do not have autism (Hayes, 1987; Kanner, 1943; NSAC, 1978).
In visual search task experiments, the participants indicate whether they see the presence or absence of a hidden target among distractors (O’Riordan, Plaisted, Driver, & Baron-Cohen, 2001). The rate of the search depends on how similar the target is to the distracters in the background. In many of these experiments, autistic individuals perform better than people that do not have autism. In a perceptual learning experiment, children with and without autism performed two visual search tasks in which a target letter was hidden among several distracter letters. Letters differed in terms of color, shape, and size. The children with autism were actually faster at detecting the letter than the children without autism. This suggests that autistic individuals may have superior visuospatial skills (O’Riordan, Plaisted, Driver, & Baron-Cohen 2001).

In another perceptual learning experiment, people with and without autism were shown seven red and yellow circles on a blue rectangular background. The circles differed in pattern and position. The autistic group performed better than the control group at detecting changes, which means they show superior discrimination (Plaisted, O’Riordan, & Baron-Cohen, 1998).
Although people with autism have difficulty shifting attention from one stimuli to another (Courchesne, Townsend, Ashoomoff, Yeung-Courceshesne, et al., 1994, Courchesne, Townsend, Ashoomoff, Saitoh, et al., 1994), they are better able to discriminate between stimuli that is new and highly similar and are better at visual search tasks. This is because people with autism process similar characteristics poorly (reduced generalization) and unique features very well. The more generalization between stimuli, the harder it is to tell them apart or discriminate. This is called the perceptual learning effect (Plaisted, O’Riordan, & Baron-Cohen, 1998).

According to the perceptual learning effect, generalization is contingent upon how well a person recognizes common characteristics. Several hypotheses attempt to explain why there is a perceptual learning effect in individuals with autism.
The overselectivity hypothesis (Lovaas, Koegel, & Schreibman, 1979) states that the transfer of learning to new environments by autistic individuals is reduced because the surrounding stimuli that was present in the original learning environment is not
present in the transfer circumstances. This would be akin to the taking of a test in a different classroom than the original one where the learning took place. This is called poor transfer (Plaisted, O’Riordan, & Baron-Cohen, 1998).

The weak central coherence theory also attempts to explain why people with autism have a superior ability to discriminate. Weak central coherence means that people with autism may focus on certain details of a stimulus rather than the stimulus as a whole thus they may perceive a highly similar stimulus as a new stimulus (Frith, 1989; Frith & Happe`, 1994).

Although people with autism are prone to impaired perception, the fact that individuals with autism performed better in visual search tasks and discrimination tasks challenges the idea of reduced shifting of attention. It is also shown that novel stimulus detection is enhanced in autism, which may predict that in any situation that requires one to respond to unique cues, autistic people will perform better than normally developing people. Based on these findings, I hypothesize that autistic individuals are less vulnerable to change blindness than individuals that do not have autism.

In my experiment, I used the method of implicit change detection in that participants did not know any changes were going to take place. The participants were twenty-five high-functioning autistic adults (Group 1) and a control group of twenty-five non-autistic adults (Group 2). The adults in Group 1 had been diagnosed with autism by clinicians according to criteria in the DSM-IV.

Each participant was shown a video of a playground scene. The video was made so that the color of the swing set changed once (from red to yellow). The change occurred right after the camera angle changed. The video was shown on a big-screen television in DVD format in a controlled classroom setting.

After the video was shown, each participant was asked if they noticed anything unusual about the playground they just saw. Next they were asked if anything changed, and if so, what? Last, they were asked if they noticed that the colors of the swing set changed. Answers were counted as correct if the participants could identify what the change was. After the experiment, the participants were debriefed and dismissed.

Results of the experiment were measured using a 2-way chi-square test. More people in Group 1 (73%, 18 out of 15) noticed changes than in Group 2 (40%, 10 out of 25), where X (1, N = 50) = 5.195, p < .05, �µ = .o5, and the critical X value = 3.84. Since the results, 5.195, were above the critical value, 3.84, the null hypothesis was rejected.

Overall, individuals with autism performed better than the individuals without autism. However, extraneous variables such as age and gender were not accounted for. This limitation makes it impossible to say for certain whether it was the condition of autism alone that produced the superior performance, because the ages and genders of the participants might have played a role in their ability as well. While the results look promising, future research will determine for sure whether autism influences change blindness.

Leave a Reply

Your email address will not be published. Required fields are marked *

9 × = forty five