CrimePsych

Abstract
The purpose of this study was to examine differences in change detection between viewings of a short edited film as well as the affect of sound on these change detections. A total of 306 participants were randomly assigned to one of two conditions, one with sound and one without sound, where they watched a short edited film on two consecutive occasions approximately five minutes apart. The results support previous research in the field of change blindness with significant results obtained for both of the conditions within this study. Not only do participants fail to notice large changes to centre of attention as well as arbitrarily placed objects, unless told to watch for changes, but also whether or not sound is present has an affect on participants ability to detect said changes. In this case, participants without the presence of sound detected more changes than those with sound.

As far back as the 1970’s researchers have been conducting systematic studies on complex areas of human perception including our visual abilities. Early research focused on eye movements in general and how they react to certain situations and stimuli but also on change blindness and the reasons behind it.

Initially change blindness explorations focused on text and simple pictures, but more recently studies have extended towards complex scenes and representations. A number of these studies have used the saccade as their prominent time to make alterations to the scenes and representations, but another method was also devised referred to as the ‘flicker task’. This involves a process where a blank slide is inserted between two slides depicting a very similar scene and participants are required to find the change made.

A number of different experiments were also conducted by Simons and Levin; in one of these, participants were randomly approached in the street by a pedestrian asking for directions, but during this interaction the pedestrian was replaced by someone else. Only half of the participants detected the change. In another study participants were required to watch a short film in which arbitrarily placed, as well as centre of attention, objects were removed or modified during angle changes or ‘cuts’. In a separate condition within this same experiment the main actor was swapped during an angle change. Again very few participants noticed the changes. Levin and Simons acknowledge that even when objects are the centre of attention we more often than not fail to detect any change.

However in this current avenue of research the affect that sound has on our ability to detect change between cuts in a film, as well as more generally our ability to simply detect change in that same film, is examined.

Research on the affect of sound and how it influences our ability to detect change is lacking. Thus it is hypothesised that more changes will be detected by participants on their second viewing of the short edited film than on the first, and that those who watch the film without sound will detect a larger amount of changes than those who have the distraction of sound present.

Method
Participants
Participants in this study were 306 undergraduate students at Macquarie University comprising 82 males and 224 females (M = 20.51 years, SD = 5.07).

Design
Students were randomly assigned to one of two possible conditions, based on their practical class on campus. In the first condition the experiment was run as a repeated measures, within groups design while the second condition was a repeated measures, between groups design.

Materials
Students were shown a short edited film (Simons, 2003) in a classroom environment of approximately 30 students per class using a ceiling mounted data projector and wall mounted projector screen which was a maximum of 15 feet away.

Procedure
In both conditions students watched a short edited film on two separate occasions; however condition one had sound while condition two had no sound. On the first watch of each condition no instructions were given except to “Pay close attention”. Upon completion students were asked to note anything unusual that occurred in the film.

Participants were then informed that changes occurred between each scene cut and were asked to watch the film a second time noting any changes.

Upon completion of the second viewing the number of changes detected was marked by a neighbour to avoid any extraneous variables or biases.

Results
Differences between first and second viewing. Number of changes detected on the second viewing (M = 1.92, SD = 1.20) were much higher than on the first (M = 0.14, SD = 0.48). In addition to this, our t-test revealed a significant difference between the two means with our calculated t of 25.80 exceeding the critical t of 1.96 (from tables).

Differences between sound and no sound conditions. In this condition the number of changes detected was higher in the no sound group (M = 2.11, SD = 1.27) than in the sound group (M = 1.73, SD = 1.11). Again our t test revealed that the difference between our two means was a significant one; our calculated t of 2.76 exceeded the critical t of 1.96 (from tables).

Discussion
The aforementioned statistics support the overall hypotheses as the null hypotheses, that there is no difference between the means, were rejected on both occasions. The number of changes detected was much larger in the second viewing than in the first, however even after having watched the edited film once, participants in the no sound condition detected more changes than the people in the sound condition.

This supports the idea that even when participants have heard a conversation before, which was the case in this study, they are still distracted by the presence of sound. The results for the first viewing versus the second viewing also fortify other research into change blindness; people have a habit of failing to detect considerably significant changes to scenes, even when they occur to centre of attention objects, as was the case in Levin and Simons experiments as well as others.

The reasons for peoples inability to detect these changes is linked not only to our attentiveness but on a much deeper level our evolution. Our visual processing systems have evolved in such a way that it can take advantage of the structure of natural scenes as well as by using bodily motion and locomotion to take cognitive shortcuts and reduce load on a person’s cognitive system.

However the current study appears to have a number of limitations, slight as they are, that may have influenced the results. Firstly, even with instructions to not talk about the research being conducted with friends during the data collection phase there is no guarantee that a lot of participants followed these instructions. Secondly, the viewing conditions for the edited film differed from classroom to classroom and were not kept constant; some participants were sitting directly in front of the projector screen while others were off to one side and various other scenarios. Thirdly, time of day could play a role in number of changes detected in that as the day drags on people get more tired and less attentive and are less likely to detect the changes.

It would be difficult to combat all of these problems as people are always going to talk to their friends, and it is unlikely that all the data gathering can be conducted at an exact time of day, however to control environmental pressures it would be advisable to at least control the viewing angle and distance from the film being watched as research has suggested that the size of the presentation being watched can influence detection ability.

Further research should attempt to investigate if sound alone plays a role in change detection i.e. have participants watch an edited film with changes such as the one used in this study but have completely irrelevant sound playing over the top of it. It has been accepted that for us to notice change we need to be attentive to the object that has changed. Whether the lack of change detection is due to sound in general or the fact that subjects are paying more attention to the conversation in said film needs to be confirmed.

In closing, the detection of changes in real life scenarios as well as complex and simple representations appears to be influenced by a number of different factors, one of which is sound. Only with further research on slightly different tangents will we discover how large a role sound plays on our attentiveness and ability to detect changes as a whole.