Interference in Visual Working Memory is Related to Less Accurate Response

Introduction

Memory is a tool on which many people rely heavily every day. How and what is remembered plays a significant role in determining how people act in their daily lives (Araya, Ekehammar & Akrami, 2003). It is important to understand how memory works as a way of understanding more about people in general, and specifically about how the brain stores information. With this information people can be helped to expand and build memory, those with learning disabilities can be treated, and eyewitness testimony can be filtered for accuracy (Dysart, Lindsay, Hammond & Dupuis, 2001). Understanding memory has and will continue to influence many disciplines and help uncover the myriad mysteries of the mind (McNamara & Wong, 2003).


Researchers have spent decades studying the mechanisms of memory (Ho, Cheung, & Chan 2003). There are several different types of memory, such as episodic, which allows people to remember specific events with striking clarity. Implicit memories are pieces of information people know, but are not aware of the source. Conversely, explicit memories are pieces of information people can remember and they can also remember when and how they acquired that information.

Most researchers believe that a piece of knowledge must pass through a series of “gates” before it is permanently stored in memory (Ward & Loftus, 1985). The first of these gates is called working memory, where objects that are being attended to at the present moment are stored. Objects in working memory can remain there anywhere from two seconds to several minutes before either being forgotten or passing though the next gate, known as short term memory.

Short term memory is similar to working memory, and not all researchers agree on the distinction between working memory and short term memory (Hinson, Jameson & Whitney, 2003) However, short term memory is considered to be the staging area where the fate of a piece of information is decided; it will either fail to be encoded as a memory, and so be forgotten, or it will pass into permastore, where long term memories are held. Approximately 7 items can be stored in short term memory at one time.

With this knowledge, researchers have been able to test memory skills under a variety of conditions and in many circumstances, to arrive at a better understanding of memory skills and function. In order to attain this knowledge, most of the research in memory has been aimed at short term or working memory (Hinson, Jameson & Whitney, 2003). For example, researchers know that information to which people are exposed in passing can sometimes remain, subconsciously, in memory and can influence future decisions (Bushman & Bonacci, 2002). This information is frequently used in advertising, where advertisers hope that frequent exposure to their products will prompt more purchases at the store. Extensive market research has shown that consumers are indeed influenced by this exposure (Bushman & Bonacci, 2002).

In the 2003 study by Hinson, Jameson and Whitney, it was found that individual differences in working memory are related to decision making that favors short-term over long-term consequences. People with more efficient working memories were better able to process and weigh the differences between a short-term reward, which appeared better at first glance, and a long-term reward, which was actually the better choice. There were some differences noted in the working memory processes of the individuals who chose the long-term reward.

One of the most popular areas of memory research is that of eyewitness testimony and identification (Dysart, Lindsay, Hammond & Dupuis, 2001). In a study by Dysart and colleagues (2001), there was significant evidence to show that eyewitnesses who had been shown mug shots of suspects prior to participating in a lineup were more likely to choose a suspect to whom they had been previously exposed. The implications of this research are monumental: A suspect can be incarcerated on the basis of testimony that may have inadvertently been planted in memory by the viewing of mug shots. With this research, the lives of many innocent people may be spared.

Another study on eyewitness performance by Ward and Loftus (1985) showed that when two people witness the same event, their memories will be different. They also found that people with certain personality traits were more likely to me misled by false information.

Verbal and visual memory are two more types of memory. Verbal refers to remembering associated with words, like how many words a person can remember when they hear them read aloud. Visual memory refers to the amount and accuracy of recollections having to do with visual stimuli, such as the details of the slide show in the current study. It has been found that verbal memory is affected by interference in working memory (Woodman, Vogel & Luck, 2001). For example, if a person is asked to remember a series of words that are read aloud, and then must complete a task that fills their working memory, they are much less likely to accurately recall the words. Since verbal and visual memory are closely related, logic would assume that the same rule would hold true for visual memory.

To this end, it is hypothesized that the more interference that is present between the stimulus and the desired response, the less accurate the participants’ responses will be. Group 1, which has only a time delay, should have the most accurate responses, and there should be a significant difference between the accuracy of the responses of group 2, which had the easy task interference condition, and group 3, which had the difficult task interference condition.

Method

Participants

Participants were 17 undergraduate students enrolled in the Thursday lab section of experimental psychology at Rutgers University. Group 1, the control group, consisted of 5 students, while groups 2 and 3 consisted of 6 students. Students were randomly assigned to groups as they entered the classroom. Students who participated were given class credit for their participation in this and several other experiments.

Procedure

It should be noted that this study was conducted after a very short study in this same lab period. Participants may have had less interest in this study because they were told it was the longer of the two, or they may have been impatient to leave. Participants were shown a brief slide show of a man visiting several stores, examining merchandise, and shoplifting several items. Group 1, the control group, was then required to sit quietly for 5 minutes, while group 2 completed an easy task and group 3 completed a difficult task. The easy task was to look for pictures of items in a newspaper and to record the page number on which they were found on a questionnaire. The difficult task was to complete the crossword puzzle in the newspaper and writing a brief assessment of one’s puzzle solving abilities.

After the 5 minute period, all participants read a narrative that told the same basic story as the slide show, but with some misleading and some reinforcing statements.

After the narrative, another 5 minute period followed in which group 1 was required to sit quietly, group 2 completed an easy task, and group 3 completed a difficult task. The easy task was again identifying pictures in a newspaper, while the difficult task consisted of answering questions based on information embedded in different newspaper articles.

After the second 5 minute period, all participants were tested on their memory of specific details from the slide show.

Results


The data indicated support for the hypothesis that the group with the most interference between the slide show and the questions would yield the most inaccurate responses. However, group 2 (easy task), out-performed both group 1 (control group), and group 3 (hard task). It appears that when slightly challenged, participants perform at their optimal level.

Group 1 performed better than Group 3, but not as well as Group 2. This may be attributable to the fact that the participants of Group 1 had nothing with which to occupy their minds except their own thoughts, and they may have grown bored and disinterested during the two periods of down time. Group 3 did have the fewest accurate responses, due to the fact that their tasks were the most challenging and required the most abstract thought. These results are in keeping with the hypothesis that, with too much interference, the brain is unable to store a lot of information in working memory. However, the hypothesis did not anticipate that no interference at all may in fact prove more distracting than a small amount of interference.

Though our experiment yielded results which seem to be an accurate demonstration of the capacities of memory, further research is highly suggested. Due to the fact that this lab class has only minimal students and several were absent during the experiment, the results may not be as reliable as possible. Having a large sample group is an important factor for any experiment. Because these experimental groups were so small, it would be advisable to try and re-create these results with a larger, more diverse subject pool.

In the questionnaire regarding the slide show, the questions were classified as reinforced: items in the slide show were mentioned correctly in the narrative (e.g. yellow candle); misled: items in the slide show were mentioned incorrectly in the narrative (e.g. white candle); slide only: items in the slide show were not mentioned in the narrative; and neutral: items in the slide show were mentioned in the slide show, but with no modifiers (e.g. candle). There were four (4) misleading questions, eight (8) slide only questions, four (4) neutral questions, and four (4) reinforced questions.

All groups performed the same or better with reinforced questions than with any other questions. Group 1, the control group, had a 90% overall accuracy for reinforced questions, 87.5% for slide only, 85% for neutral questions, and only 40% for misleading questions. Group 2, the easy task group, had 87.5% overall accuracy for both the reinforced and slide only questions, 79.2% for neutral questions, and 66.7% for misleading questions. Group 3, the hard task group, had 95.8% overall accuracy for reinforced questions, 85. 4% for slide only questions, 70.8% for neutral questions, and only 33.3% for misleading questions. Out of all the groups, group 3 had the best overall accuracy for reinforced questions, and the worst overall accuracy for misleading questions.

Group 2 had the least variance, a 20.8% difference, in their overall accuracy between the misleading and reinforced questions, group 1 had a 40% variance, and group 3 had 62.5%, the most variance.

All 3 groups rated themselves similarly to one another on confidence judgments, and across all categories of questions. Across all 20 questions, Group 2 rated themselves an overall confidence of 87.81%, Group 1 rated themselves an overall confidence of 88.48%, and Group 3 rated themselves an overall confidence of 88.57%. All groups rated their performance within 1% of each other. However, Group 3 (hard task), who yielded the least accurate responses, rated themselves highest in confidence. Group 2, (easy task), who yielded the most accurate responses, rated themselves lowest in confidence, and Group 1 (control group) rated themselves right in the middle.

Group 3 showed an overall accuracy for confidence judgments of 71.35%, Group 1 showed 75.63%, and Group 2 showed 80.21%. Group 2 did nearly 10% better than group 3 and nearly 5% better than group 1.

Conclusion

A most interesting phenomenon lies in the results of the confidence judgments. It appears that participants have no true idea of their actual accuracy in answering questions. Regardless of their actual accuracy, all 3 groups rated themselves fairly high on their confidence levels, and extremely close to one another. These results are in keeping with the findings of Elizabeth Loftus (1986), who demonstrated that individuals’ confidence of their answers has no bearing on their actual results.

It appears that, for working memory to function at an optimal level, some interference or distraction is desirable. When individuals are forced to concentrate too avidly on a task, they appear to lose interest and become easily distracted from their goal. When there is too much interference or distraction, however, the individual is unable to recall all the information from working or short term memory, probably due to the fact that they are forced to channel their concentration almost completely into other areas. Therefore, the data lead to the conclusion that a low level of distraction has the capability of driving the memory to achieve greater results.

It would be interesting to try to replicate these results with a larger group of participants, and also to create more distinction between the difficulties of each group’s task. Doing so may yield more detailed information on the intricacies of working memory.

Leave a Reply

Your email address will not be published.