Weapon Focus and Inattentional Blindness

Published Date: 23 Mar 2015 Last Modified: 31 May 2017

Keywords: eyewitness memory accuracy, weapon focus effects

The aim of this research is to investigate the effects of weapon focus and inattentional blindness on eyewitness memory. 91 participants were randomly assigned to one of six conditions. Participants performed one of three task; count the number of bags (consistent attentional set), count the number of people in white coloured tops (inconsistent attentional set), or count nothing (control), while watching a video depicting a theft consisting of presence or absence of weapon. Participants then completed a questionnaire. The results obtained reveal that there was a main effect of weapon focus on eyewitness memory. However, no significant results were found for the main effect of inattentional blindness and the interaction effect of weapon focus and inattentional blindness on eyewitness memory.

The Effects of Weapon Focus and Inattentional Blindness on Eyewitness Memory

Despite the advances in technology in forensic sciences, eyewitness testimony is still widely used as evidence in court trials to both recall the events occurred and to identify the perpetrator. Evidently, eyewitness testimony has a crucial impact on juror's decision making. However, unintentional errors concerning identification of the perpetrator tends to occur. Therefore, several studies have been conducted to study the factors that may impair eyewitness memory. The current study aimed to investigate the effects of weapon focus and inattentional blindness on eyewitness memory as doing so may yield results that are more applicable to real-life situations whereby witnesses are not completely attentive to the event that occurs or on the perpetrator, hence affecting their memory for it.

Crimes involving weapons are on the rise. There has been conflicting arguments in this field of research as some studies suggested that violent conditions such as weapons is associated with emotional arousal that may benefit memory, however, experts in the area favoured the notion that scenarios involving the presence of weapons negative affects eyewitness performance. This phenomenon is known as the weapon focus effect. Weapon focus effect is defined as the attention focused on a weapon in an event which in turn results in the reduction of attention directed towards other details (Saunders).

According to Easterbrook (1959), weapon focus effect can be explained using cue-utilisation theory which states that upon detecting the presence of a weapon, witnesses experience increased anxiety which will then restrict and focus their attention to the weapon instead of the perpetrator (Easterbrook). Weapons illicit high levels of stress and emotions as weapons are perceived as dangerous. Thus, according to Easterbrook, the heightened emotional arousal leads to a decrease in attentional capacity; a decrease in the variety of cues from the stimulus (weapon) and its environment to which the individual is present in. Specifically, an individual's attention will be restricted and focused on the central information of the stimulus (weapon) at the expense of peripheral; minor, information (perpetrator's features). Therefore, information or cues central to the source of the emotional arousal (weapon) will be encoded while peripheral information or cues will not.

Or does this sound better? Or should I rewrite it based on both contents?

As arousal increases, the number of perceptual cues utilized decreases. This reduction begins with peripheral cues at a lower level of arousal and later if arousal increases it will affect central cues. At an optimum arousal level, when utilization of peripheral cues is minimized allowing total attention to be paid to central cues, performance on a central task could actually improve. In accord with this theory, it follows that weapon focus effect occurs because in a crime situation, the weapon becomes the central cue the criminals' characteristics becomes a peripheral cues and as arousal increases encoding of the peripheral cues decreases. Perceptual narrowing occurs, focus on the gun instead of perpetrators' characteristics

It is argued that witnesses would consider weapons as the central cue because witnesses would consider information pertaining to questions such as, 'Is it about to be utilized?' and 'Is it aimed at me?' (Kramer et al., 1990, as cited in Pickel, ). This view has also been supported by Christianson (1992) who found that emotional arousal leads to narrowing of attention such that attention is directed to central details, at the expense of peripheral information. This is supported by studies that measures eye fixation patterns that reflects visual attention. Studies in this area suggests that emotional scenes elicit longer fixation durations on central cues. It was also found that central details of the scenes were better retained as compared to peripheral details.

Loftus, Loftus, and Messo (1987) demonstrated a classic example of the weapon focus effect. In this experiment, 36 participants viewed a series of 18 slides depicting customers ordering food at a fast-food outlet. Participants in the controlled group viewed a customer handing the cashier a cheque for the purchase made followed by the cashier returning some change. Meanwhile, participants in the experimental group viewed a customer pulling out a gun followed by the cashier handing the customer some money. Participants' eye movements were recorded using a corneal reflection device that delivers a television picture of the scenes observed by the participants combined with a spot light that moves in accordance with participants' eye movements. Participants were also asked to answer a short questionnaire and to identify the target man from a selection of 12 photographs. Results from this experiment revealed that participants in the weapon condition spent a considerable large amount of time fixating on the gun as compared to participants in the control condition; did not fixate much attention on the check. Results from the memory questionnaire and line up test show that participants in the weapon condition scored lower than participants in the control condition. Therefore, results from the eye movement data indicates that weapons distract participant's attention from other important cues thus affecting their ability to identify the perpetrator. However, although significant results were found, one should be cautious with the results as participants in both conditions were exposed to relatively two different scenarios whereby not only was there a manipulation of the presence and absence of weapons, there was also a manipulation of presence and absence of a violent crime. Therefore, the dependent variable may have not likely measured the same thing.

Easterbrook's cue-utilisation hypothesis has also been supported by an abundance of literature in the area. For instance, Pickel, Ross, & Truelove (2006) looked at whether weapons have an effect in capturing attention. 230 participants were randomly assigned into one of four conditions whereby they were told they would be watching a staged scene. Participants in the experimental groups were given a talk on the weapon focus effect while participants in the control groups were given a talk on eyewitness confidence. A minute after the lecture, an actor entered the classroom demanding to see a professor who had awarded him a failing grade while holding either a book (neutral object) or a gun (weapon). Participants then completed two forms; Thayer's (1989) Activation-Deactivation Adjective Check List Short Form and a questionnaire testing participants' memory of the scene. Results revealed that participants who were introduced to a neutral object scored better in the memory test as compared to the participants who were exposed to a weapon which provides support for Easterbrook's cue-utilisation hypothesis (Pickel et al.). This study further found that participants exposed to the lecture on the effect of weapon focus had a better memory score as compared to participants exposed to the lecture on eyewitness creditability. This finding suggests that the weapon focus effect can be overcome if it is made known to individuals.

Besides that, Hope & Wright (2007) examined the effects of weapon on attention. Participants in this experiment viewed a slideshow consisting of 13 pictures of a simulated event of a man in a grocery store. In one of the slides, the perpetrator is seen withdrawing either a gun (weapon condition), a colourful feather duster (unusual condition) or a wallet (control condition). Participants then completed a 20 minute filler task before completing a questionnaire that measured the memory of the event. Data collected revealed that participants exposed to the weapon condition had a lower accuracy regarding the perpetrator's physical characteristics but had a higher accuracy in recognizing the weapon (Hope & Wright). This finding suggests that individuals tend to fixate their attention on weapons instead of the perpetrator and thus affecting their ability to identify the perpetrator which is crucial in courtrooms.

In addition to Hope & Wright's results, Pickel (2009) investigated the effects of weapon on memory. In this experiment, 127 participants were shown a video depicting a perpetrator robbing two victims; a male and female, while holding either CD (neutral object) or a gun (weapon). Upon watching the video, participants completed a questionnaire which tested their memory of the perpetrator's physical appearance. Results obtained show that participants in the weapon condition scored lower in recalling the perpetrator's physical appearance (Pickel). This finding also supports the cue-utilisation hypothesis introduced by Easterbrook (1959) and provides evidence that weapons may impair an individual's ability to recall the eyewitness event.

Maass and Kohnken (1989) conducted a study to investigate the weapon focus effect on witness recall and recognition. Their experiment differed from other research in the area that used pictures or movies to depict an eyewitness event as they used a syringe as a weapon simulation that is perceived as harm to participants. 86 students were recruited for this study. A false questionnaire regarding mood and health behaviour was first administered to the subjects. Upon completion, subjects were led into a room whereby a female confederate approached them while holding either a syringe partially filled with a yellow liquid or a pen. Participants then completed the same mood scale and a couple of filler tasks for 20 minutes. Following that, participants completed a recognition and recall task. The recognition task involved identifying the target (female confederate) from a lineup of seven people while the recall task involved answering a questionnaire that accessed participants' recall of target's facial features. Results from the study revealed that participants exposed to the syringe performed more poorly on the recall and recognition task as compared to participants exposed to the pen.

Moreover, Steblay (1992) conducted a meta-analysis review on several weapon focus effect studies. 19 studies were reviewed that hypothesised that weapons negative affects witness's ability in identifying criminals. Results from the meta-analysis revealed that there is a difference in presence and absence of weapon conditions, with presence of weapons in a crime leading to a reduced accuracy in indentifying perpetrators of a crime.

In a recent study by Mitchell, Livosky, and Mather (2011), the findings of the effect of weapons on eyewitness memory was replicated. 83 participants were randomly assigned into one of four conditions whereby each of them viewed a video of a business exchange between two males whereby one of them is seen pulling out either his bare hand (control), a stalk of celery (unusual), or a handgun (weapon) from a briefcase. Shortly after, the participants completed a memory questionnaire. Results obtained reveal a classic weapon focus effect whereby participants in the weapon conditioned scored lower than participants in the control and unusual condition.

Besides exploring the effects of weapon focus, this research also aimed to look at inattentional blindness which was first coined by Mack & Rock (1998). They defined inattentional blindness as a failure to perceive an unexpected stimulus as attention is devoted on another task or object (Mack & Rock). Individuals tend to fail to perceive everything detail of the environment at all times, even though there appears to be no apparent factors that hampers their vision. In some situations, the consequences are trivial. For instance, Simon and Chabris (1999) found that individuals who were focused on finding their seats in a crowded cinema often did not notice their friends waving their arms at them, trying to obtain their attention. However, sometimes the consequences may be deadly. For instance, in an experiment, pilots were asked to operate a flight simulator whereby the flight console instruction was projected onto the windshield of the cockpit. It was assumed that this would have decreased errors as the pilots would be viewing both the console information as well as the external world at the same time. The results from this study was truly shocking. Some of the pilots tried to land the airplane as requested by the console information although another airplane was obstructing the runway. When questioned, the pilots reported not being aware of the obstruction; they did see an airplane on the runway despite looking at it directly.

To better understand the factors that are involved in focusing attention to an unexpected stimulus, Neisser (1979, as cited in Rivordo et al., 2011) introduced a perceptual cycle framework which accounted for attention capture and capture of awareness. Typically, Neisser studied stimulus properties that would influence the likelihood of an individual noticing an unexpected object or event (bottom-up processing) as well as the processes that is under the control of the observer which influenced what they notice (top-down processing). Thus, through his efforts, it has been found that unexpected stimulus alone and distinctiveness of the unexpected stimulus do not capture awareness. Further research in the area has found that a stimulus is more likely to capture awareness if it is consistent with the attentional set. An attentional set is defined as devoting attention to a particular characteristic of a stimuli (Rivordo et al., 2011). Therefore, the likelihood of noticing the unexpected stimulus increases if it shares similar perceptual features with the primary task. It has been found that when individuals adopt a attentional set, they adopt a top-down processing which overrides the individual's ability to capture other available information. In a typical inattentional blindness study, an attentional set is established by providing instructions for a primary task. In regards to the current study, the primary task will be counting a stimulus while the secondary task will be watching the video shown.

Neisser's theory was put to test by Most et al. (2001) that is based on a study by Mack and Rock (1998). In their study, 128 participants were recruited. Participants were exposed to a computer programme whereby four white and four black T and L shapes moved independently and randomly. As the objects moved, they would collide with one another and bounce off the edges of the screen. Participants were asked to fixate their attention to a central point in the middle of the screen but were asked to keep tally of the number of times the objects (either white or black T and L shapes) bounced off the edges of the screen. Participants completed five trials. The first and second trial did not introduce any unexpected stimulus. In the third trial, a cross shape, either white or black in colour appeared from the left of the screen and moved horizontally towards to the right of the screen. Participants were then given a questionnaire that asked if they had seen anything unusual and provide details of the object. Following that, participants completed a fourth trial whereby the cross appeared once again. They then completed the same questionnaire. In the fifth trial, participants were asked to fixate their attention on the centre point of the screen and not to keep track of the bounces (full attention). Finally, they completed the same questionnaire. Results revealed that when the unexpected object (cross) was similar to the attended objects, a greater number of participants noticed the stimulus. Specifically, 94 percent of participants who attended to the white coloured shapes noticed the cross when it was white while only six percent of participants in the same condition noticed the cross when it was black. Similar findings were found for participants who attended to the black coloured shapes; 94 percent noticed when the cross was black, while none of the participants noticed when it was white in colour (Most et al.). In a second part to their study, 32 participants were recruited. The materials and procedures used were identical to those in the first experiment. However, instead of attending to T and L shapes, circles and squares were used and instead of a black cross, the unexpected object was a red cross. Thus, it differed in shape and colour. Results from this study reveal that about 30 percent of participants failed to notice the unexpected object. Therefore, results from both studies supports Neisser's perceptual cycle framework that states that distinctiveness alone cannot account for attentional capture, instead, it is the similarity between the unexpected stimulus to other objects present that influences attentional capture.

Once Neisser's theory has been established and supported, Simon & Chabris (1999) conducted an experiment to examine the role of an attentional set in inattentional blindness. 228 participants were recruited for this study. They were asked to watch a video of two teams consisting of three players playing a game of basketball; one team in white tshirts, another team in black tshirt. Participants were told keep tally of the number of passes made by either the white or black team. Halfway through the video, participants were exposed to either one of two unexpected events. In one condition, a woman holding carrying an umbrella walked from the left corner of the area to the right and disappearing off screen. Meanwhile, in the second condition, a woman in a gorilla costume is seen walking in the same direction; from the left to the right. After the video was played, participants were asked to write down their counts on a piece of paper followed by completing a surprise questionnaire that questioned if they had noticed anything unusual. Overall, it was found that only 54 percent of participants noticed the unexpected event. Specifically, the results revealed that participants who counted the number of passes of the black team (primary task) noticed the black Gorilla more; (consistent attentional set in regards to colour of team to colour of the Gorilla costume) while participants who were asked to observe the actions of the white team hardly noticed the black Gorilla (inconsistent attentional set). In contrast, there was little difference in participants who were exposed to the woman with an umbrella. This could be attributed to the fact that the Gorilla was black in colour whereas the lady with an umbrella was dressed in pale colours. Thus, their findings reveal individuals are more likely to notice an unexpected stimulus if it shares similar features, in this case, colour of the gorilla with the colour of the team (Simon & Chabris).

Besides that, Hyman et al. (2010) examined the effects of inattentional blindness while walking. 151 individuals were observed by two observers who were positioned at both ends of a walkway. Out of the 151 participants, 78 individuals were without any electronic device, 24 individuals were using their phones, 27 individuals were using their music player while the remaining 22 individuals were walking in pairs and engaging in a conversation. A brightly coloured dressed unicycling clown was placed somewhere in the middle of the walkway. At the end of the walkway, participants were asked if they had seen anything unusual and to describe it if answered yes. Participants who answered no were specifically asked if they had seen the unicycling clown. Results from their study revealed that participants who were on their cell phones were less likely to notice an unexpected stimulus; a unicycling clown, while walking. Specifically, 75 percent of individuals were inattentionally blind to the unicycling clown while 51 percent of individuals without electronic device, 61 percent of individuals with a music player, and 71 percent of individuals in pairs noticed the unicycling clown (Hyman et al.). This finding reveals that an unexpected stimulus can pass an individual's visual field if attention is focused on another task. Although it could be argued that the cell phone users were less likely to pay attention to their surroundings, Strayer et al., (2003, as cited in Hymen et al., ) found that mobile phone users were as likely to look at objects while in a driving simulation, but were less likely to remember the objects when compared to individuals not engaged in another task. Moreover, although this study did not study the influence of attentional sets in capture of attention, it provides substantial evidence for the occurrence of inattentional blindness in daily life.

Furthermore, Karns & Rivardo (2010) investigated the role of attentional set in inattentional blindness. 86 participants were randomly assigned to one of two scenarios (family emergency or restraining order) and also to one of two unexpected situations (family confrontation or gorilla). Participants were exposed to 4 videos that were shown on a single screen. The top left and bottom right quadrant contained distracter videos of the lobby of a dormitory. The top right quadrant contained a video of a target person walking through the student lounge while the bottom left quadrant contained the unexpected event. Every participant was exposed to the same videos except for the unexpected event. Half of the participants were exposed to a confrontation between two individuals near a staircase while the remaining half of the participants were exposed a man clad in a gorilla costume walking down the same staircase. Participants in the restraining order scenario were provided with a picture and physical description pertaining to a target individual who has a restraining order against him and was no longer allowed to be on campus. On the other hand, participants in the family emergency scenario were provided with the same picture and physical description. However, they were informed that the target's family was looking for him due to a emergency. Following that, participants were informed to act a security officers and be on the lookout for the target person. They participants were instructed to pause the video upon identifying the target individual. Finally, participants were asked to completed a questionnaire that inquired if they had seen anything unusual on campus. The results obtained shows that consistent attentional set increases the probability of noticing the unexpected stimulus; participants in the family emergency condition were more likely to notice the gorilla (71 percent) than the confrontation scene (33 percent) while participants in the restraining order condition was more likely to notice the confrontation (77 percent) rather than the gorilla (40 percent) (Karns & Rivardo). This finding reveals that an attentional set that is consistent with the content of the event will reduce inattentional blindness.

A recent study by Chabris, Weinberger, Fontaine and Simons (2011) simulated a incident that occurred in Boston whereby a policemen on a chase ran past an assault but claimed to not have noticed it. This study was a first attempt to study inattentional blindness in real-world events. In the first part of the experiment, 20 students were recruited and tested individually. Each participant was asked to chase a male confederate for 400 meters during night time. Participants were asked to maintain a distance of at least 9 meters while keeping count of the number of times the confederate touched his head. About 125 meters into the chase, about eight meters away from the two runners, two male confederates were beating up another male confederate. At the end of the chase, the experimenters asked the participants if they had witness anything unusual along their pursuit. Results from their study revealed that only 35 percent of participants noticed the fight. This findings suggests that inattentional blindness is applicable to real life situations. However, one could argue that it was due to the darkness that participants failed to notice the fight. Therefore, the experimenters replicated the study by repeated the same procedure but held the chase during daytime on. Results from the second study reveal that only 56 percent of participants noticed the fight, thus providing evidence that inattentional blindness still persists. Once again, although this study did not integrate the effects of attentional sets in noticing an unexpected stimulus, compared to the study by ?, this study manipulated the participants' task (asked to count the number of touches to the head). Meanwhile, the study by ? merely manipulated the unexpected stimulus while the primary task of participants were not manipulated.

Another study by Lane (2006) investigated the effects of engaging in two task simultaneously on memory. In his study, 144 participants were randomly assigned into two groups. Participants in both groups were asked to watch a slideshow of a scenario whereby a man is seen entering a office, repairs a chair and finally stealing some money and a calculator. Participants in the first group were asked to watch the slideshow while listening to some songs. They were further instructed that the tape would be stopped at any given time, and when that happens, they were to state the title or artist of the last two songs. Following that, participants completed a music recognition test. Then, participants completed a word search puzzle before answering a memory questionnaire. Meanwhile, participants in the second group were asked to first view the slideshow before listening to the songs. Following that, participants completed the memory questionnaire before completing the music recognition task. Results from this study disclosed that participants who were asked to participate in two task simultaneously displayed poorer memory scores for the eyewitness event than participants who completed both task separately.

Although there has been extensive research on both inattentional blindness and eyewitness memory separately, there appears to be little integration between them. To date, there is only one study that integrated both these areas. Rivardo et al. (2011) conducted a study to examine how attentional set is able to influence inattentional blindness for a simulated eyewitness incident and the consequent memory for it. 187 students were randomly assigned into one of three attentional sets condition; consistent attention set (count the number of individuals with and without shopping bags), inconsistent attentional set (count the number of individuals in wearing and not wearing blue shirts), and control condition (count nothing). The attentional set in this experiment was manipulate according to the similarity between the primary task (what to count) and the critical item (shopping bag stolen). At the start of the experiment, participants were given a word list to memorize. After that, a video of a theft of a shopping bag that occurred in a mall was shown. While the video was shown, participants completed the task they were assigned to according to the attentional set conditions they were in. After the video was shown, participants wrote down their tallies on a piece of paper followed by writing down as many of the words they remembered from the word list. Finally, participants completed a questionnaire based on the video. Results from this study revealed that participants in the consistent attentional set condition (count the number of bags) were more likely to notice the theft of a shopping bag as compared to participants in the inconsistent attentional set (count number of shoppers in blue t-shirts). Subsequently, participants who were inattentionally blind to the theft showed low accuracy in reporting details of the theft (Rivordo et al.).

Considering the studies conducted on both weapon focus and inattentional blindness, the current study raised the question: what are the effects of weapon focus and inattentional blindness on eyewitness memory? This area of research was chosen as little efforts have been made to incorporate these two fields of research. Moreover, the phenomena of inattentional blindness is more reflective of an actual eyewitness event as witnesses are more likely to be occupied with another activity or task such as shopping, driving, and talking to another person during the event.

There are three hypotheses. Firstly, it was predicted that participants in the weapon condition will have lower accuracy of eyewitness memory. This hypothesis is in line with research conducted by Pickel et al. (2006), Hope & Wright (2007), and Pickel (2009). Secondly, it was hypothesized that participants in the consistent attentional set condition are more likely to notice the theft compared to participants in the inconsistent attentional set condition and thus are more likely to have a better accuracy of the eyewitness event. This hypothesis is consistent with the research by Karns & Rivardo (2010) and Rivardo et al. (2011). Lastly, I hypothesised that participants exposed to the presence of a weapon combined with a consistent attentional set will produce higher accuracy regarding the eyewitness event. I hypothesized that the interaction effect will have completely opposite results in comparison to the main effect of weapon focus as a research conducted by Kane (2012) found that when weapons were present, the changes to the scenes were detected significantly faster compared to scenes with an absence of weapons and thus having an effect on memory of the event. Therefore, I hope to find similar results in this proposed research by incorporating inattentional blindness instead of change blindness. Although it can be argued that both change blindness and inattentional blindness are two different phenomenon, there appears to be one crucial similarity between the two of them. Both inattentional blindness and change blindness involves a failure in reporting a visual stimuli that is otherwise quite visible. Therefore, it has been found that focused attention is necessary for both phenomenon's to be overcome.

There are a couple of implications of conducting this study. If significant results are found, inattentional blindness may reduce the number of unreliable witnesses. This is possible as in an actual eyewitness event, witnesses are usually engaging in another task and this research attempted to create a simulated eyewitness event in a laboratory study by including a primary task. Therefore, the findings obtained can be used to improve eyewitness testimony in a courtroom. This can be done by merely asking the witnesses what they were doing while the event occurred. Upon knowing that the witnesses were concentrating on a primary task, the jury would be able to make better decisions by dismissing the accounts of eyewitnesses and focusing more on evidence. Besides that, there also appears to be a theoretical and practical implication to this study. Theoretically, further research can be conducted to investigate if both the phenomenon's of weapon focus and inattentional blindness are automatic or controlled processing. If it is a controlled processing, the practical implication that arises is that individuals who are more likely to observe a crime committed by an armed criminal, for example a bank officer, could perhaps be trained in order to prepare them to perform better as eyewitnesses.

Method

Design

This study employed a 2 x 3 between subjects design. The independent variables were weapon focus and attentional set. The first independent variable, weapon focus had two levels; presence of a weapon and absence of a weapon. The second independent variable, attentional set had three levels; consistent; count the number of bags, inconsistent; count the number of people in white coloured tops, and control; count nothing, just focus on watching the video. Consistent and inconsistent attentional set refers to the degree of similarity between the primary task; counting either bags or white coloured top, to the unexpected stimulus; bag being stolen. The dependent variable will be eyewitness memory, operationally defined as scores on the questionnaire which tests memory of the event.

Participants

180 psychology undergraduate students from a local university were recruited. The sample size was determined by conducting a power analysis using a power of 0.8 with an estimated small effect size, f = 0.25 (Faul, Erdfelder, Lang, & Buchner, 2007). A convenience sampling technique was adopted due to financial and time constraints. 89 data of participants who knew the actors in the video were discarded to avoid potential conflict of interest that would affect the validity of the results obtained.

Materials

In order to manipulate the first independent variable; weapon focus, two different videos were used. The first video depicted a scene of several customers entering a restaurant. Towards the end of the video, a male perpetrator casually walked up to the restaurant and quickly stole the bag of a male student who was seated near the entrance of the restaurant. The male perpetrator is then seen fleeing the scene. In this video, there was an absence of weapon. Meanwhile, the second video depicted a scene similar to the first video, however, instead of a man stealing a bag and running away, the man produced a knife and demanded the male customer to hand his bag over before fleeing the scene. Meanwhile, the dependent variable was measured by administering a questionnaire that contained inattentional blindness based questions regarding the content of the video as well as a recall test regarding the theft such as "Did you notice a theft?", "What was the gender of the thief?", and "Was the thief carrying any object?".

Procedure

The current research incorporated some of the procedures used in the research by Rivardo et al. (2011). In this study, participants were randomly assigned to one of six conditions. Firstly, the experimenter briefed the participants on one of the three primary attentional task; count the number of bags (consistent attentional set), count the number of people in white coloured tops (inconsistent attentional set), or count nothing (control), that is to be completed while watching the video. Next, the video was played. After the video was shown, participants were asked to complete the questionnaire. Lastly, participants were debriefed on the true nature of the experiment. A pilot study was carried out to test the procedure.

Results

After the experiment was conducted, the scores on the questionnaires were tabulated. This was performed by summing up the number of correct answers. The score achieved by each participant was out of a total of 12. The means are recorded in Table 1 and Table 2.

Table 1

Comparison of means for weapon focus

Weapon focus

N

Mean

SD

Eyewitness

Memory

Absence

Presence

46

45

8.640

3.539

0.454

0.456

Table 2

Comparison of means for inattentional blindness

Inattentional

blindness

N

Mean

SD

Eyewitness

Consistent

31

6.233

0.551

Memory

Inconsistent

Control

32

28

5.813

6.223

0.540

0.579

As shown in Table 1, participants who were exposed to the absence of weapon scored higher than those exposed to the presence of weapon. Meanwhile, as shown in Table 2, participants who were placed in the consistent attentional set condition scored the highest, followed by those in the control group, and then the participants in the inconsistent attentional set condition.

A normality test was performed to verify if the sample size is normal. The results showed that the normality assumption of skewness for weapon focus (Skewness = 0.022) and inattentional blindness (Skewness = 0.061) is met while the normality assumption of kurtosis for weapon focus (Kurtosis = -2.045) has not been met but has been met for inattentional blindness (Kurtosis = -1.467). Specifically, for weapon focus, skewness and kurtosis has not been met for absence of weapon, (Skewness = -2.083, Kurtosis = 5.687) while skewness and kurtosis has been met for presence of weapon, (Skewness = 0.236, Kurtosis = -1.651). Meanwhile, the skewness and kurtosis has been met for consistent attentional set (Skewness = -0.937, Kurtosis = -0.438), inconsistent attentional set (Skewness = -0.351, Kurtosis = -1.590), and control (Skewness = -0.622, Kurtosis = -1.067). Since the sample size was less than 2000, Shapiro-Wilk test was performed. However, the Shapiro-Wilk test is not assumed for both absence and presence of weapon, (p < 0.05). The Shapiro-Wilk test is also not assumed for consistent attentional set, inconsistent attentional set, and control, (p < 0.05). Although the normality assumption for all levels of both weapon focus and inattentional blindness has not been met, the data can still be analyzed using an independent factorial ANOVA test as it is a robust test. As such, it is not sensitive to the violation of normality (Field, 2009).

To explore the significance of the descriptive statistics results obtained above, an independent factorial ANOVA was run because of three reasons. Firstly, it is a factorial design as there are two independent variables. Besides that, the scale of measurement for both the independent variables are nominal while the scale of measurement for the dependent variable is ratio. Lastly, it is an independent design as participants only experienced one level of each independent variable.

Levene's Test of Equality of Error Variances is not assumed (p < 0.05). However, because an ANOVA test is a robust test, even though the assumption is violated, significant differences can still be found (Field, 2009). The independent factorial ANOVA revealed that there is a main effect of weapon focus on eyewitness memory, F (1, 85) = 62.869, p < 0.001. Thus, referring back to Table 1, participants in the absence of weapon condition scored significantly higher (M = 8.64, SD = 0.45) than participants in the presence of weapon condition (M = 3.54, SD = 0.46). Therefore, the first hypothesis was supported. However, there is no significant main effect of inattentional blindness on weapon focus, F (2, 85) = 0.192, p > 0.05. Therefore, the second hypothesis was not supported; participants in the consistent attentional set condition are more likely to notice the theft compared to participants in the inconsistent attentional set condition and thus are more likely to have a better accuracy of the eyewitness event. Furthermore, there is no significant interaction effect of weapon focus and inattentional blindness on eyewitness memory, F (2, 85) = 2.625, p > 0.05. Therefore, the interaction hypothesis was also not supported; participants exposed to the presence of a weapon combined with a consistent attentional set will produce higher accuracy regarding the eyewitness event.

Discussion

The results obtained show that presence of weapons in a crime reduces eyewitnesses' ability to recall details of the event accurately. This finding supports the first hypothesis which stated that participants in the presence of weapon condition will have a lower accuracy of eyewitness memory as compared to participants in the absence of weapon condition. However, the findings of this study failed to yield significant results for the main effect of inattentional blindness as well as for the interaction effect of weapon focus and inattentional blindness on eyewitness memory. Therefore, the second hypothesis was not supported which stated that

participants in the consistent attentional set condition are more likely to notice the theft compared to participants in the inconsistent attentional set condition and thus are more likely to have a better accuracy of the eyewitness event. Finally, the third hypothesis was also not supported which stated that participants exposed to the presence of a weapon combined with a consistent attentional set will produce higher accuracy regarding the eyewitness event.

The weapon focus effect was replicated in the present study which is consistent with the study conducted by Loftus, Loftus, and Messo (1987) who also found that the presence of weapons reduces the accuracy of eyewitness memory in terms of recalling details of the event and identifying the perpetrator. Their findings is further supported by an additional method employed whereby the researchers tracked the eye movements of the participants and found that participants in the weapon condition fixated more on the gun than on other objects in the environment. Although the current study did not conduct an eye-movement data collection, similar results were still obtained as both studies employed the same method of placing participants into one of two conditions: presence or absence of weapon. Thus, the similar results in the present study obtained may have followed the theory proposed by Easterbrook that stated that in the presence of weapons, the increase in anxiety restricts participants attention to the weapon and not other details in the environment.

Besides that, the results of the present study is also consistent with the study by Pickel, Ross, and Truelove (2006) who found a significant decrease in memory scores in participants who were exposed to the weapon condition. Although the researchers introduced a weapon to the participants via a staged scene while the present study used a video, similar results could be obtained because in both studies, there was a presence of threat: directly or indirectly. Thus this lead to an increase in arousal levels, which according to Easterbrook leads to a narrowing of attention whereby attention will be restricted to the weapon instead of the perpetrator.

The current study's result is also consistent with the study conducted by Hope & Wright (2007) which stated that participants exposed to the weapon condition had a lower accuracy regarding the perpetrator's physical characteristics but displayed a higher accuracy in identifying the weapon used. Although participants in Hope & Wright's study were exposed to a neutral object (a wallet) in exchange of a weapon while participants in the current study were either introduced to a weapon or no object at all, this study confirms Easterbrook's theory that weapons alone are responsible for the narrowing of attention towards the weapon.

Furthermore, the findings by Pickel (2009) also supports the present study's results. In his research, it was found that the presence of a weapon in an eyewitness event influences eyewitness's ability to recall the perpetrators physical appearance negatively. Similar results were replicated in the present study because in both the current study and the study by Pickel, both researchers introduced participants in both the absence or presence of weapon condition to a similar scene: a theft. The only difference in both the videos were the absence of presence of a weapon. Therefore, the significant results obtained can be highly attributed to the presence of a weapon.

In addition to that, the results of the current study is also consistent with the study conducted by Maass and Kohnken (1989) who found that participants exposed to an object that is perceived to be harmful perform poorly in recalling the details of the event as well as recognizing the perpetrator via line-up test. Although the study by Maass and Kohnken involved exposing participants directly to a threatening object, similar results could have been obtained as in both studies, there was a presence of threat either directly or indirectly to the participants. This threat leads to an increase in arousal levels or anxiety emotion to be specific. Thus, according to Easterbrook, this increase in arousal levels will lead to a narrowing of attention whereby attention is focused on the weapon instead of the perpetrator.

Meanwhile, the present study's results were also found to be consistent with the meta-analytic study conducted by Steblay (1992) who reviewed 19 studies on the weapon focus effect and found that presence of weapons in a crime leads to a lower accuracy in eyewitness memory. Although the approach by both the present study and Steblay's study was different, similar results were yielded because the studies reviewed by Steblay involved exposing participants to videos of crimes whereby there was either a presence or absence of weapons.

Finally, the weapon focus effect was replicated in the present study which is consistent with the study conducted by Mitchell, Livosky, and Mather (2011) who found that the presence of weapons decreased participants' ability to recall correctly the details of the eyewitness event. Similar results was obtained due to the similarity in the levels of the independent variable; weapon focus. In Mitchell, Livosky, and Mather's study, they exposed participants to either a bare hand (control) or a handgun (weapon) which is similar to the current study: no object in hand (absence of weapon) or knife (weapon). Therefore, the results obtained in both the studies revealed that weapons alone are responsible for the lower accuracy in eyewitness memory.

On the other hand, the present study failed to replicated the results obtained by Simon & Chabris (1999) with reference to the effects of attentional sets on inattentional blindness. Simon & Chabris found that consistent attentional sets aids in noticing an unexpected event. The present study was unsuccessful in attaining the same results as the unexpected event in Simon & Chabris's study was rather apparent: a black gorilla amidst a basketball game whereas the unexpected stimulus in the present study was the theft of a bag in a restaurant. Therefore, it is believed that the results were not replicated due to the differences in the context of the unexpected event.

Besides that, the current study's results were also found to be inconsistent with prior findings by Hymen et al. (2010) which concluded that if individuals were preoccupied with a task at hand, they would unlikely to notice an unexpected event despite its unusualness and clarity. The current study failed to yield any effect of inattentional blindness on eyewitness memory as the study focused more on specifically the influence of attentional sets on capture of attention while the study by Hymen et al. focused on preoccupation with another task.

The study by Karns & Rivardo (2010) also contradicts the findings of the current study as they successfully found a positive effect of consistent attentional sets on the probability of noticing an unexpected event. The present study failed to achieve similar results because in the study by Karns & Rivardo, the attentional sets were manipulated differently. Instead of asking participants to count an object that is similar or different to the unexpected stimulus, participants were provided with a scenario that was either similar or different to the unexpected stimulus. To illustrate, the results from the study found that participants who were informed to look for a target individual who has a restraining order were more likely to notice a confrontation between two individuals while participants who were informed to look for a target individual whose family was looking for him due to an emergency were more likely to notice a gorilla walking on campus.

Furthermore, the current study failed to replicate the results found by Chabris, Weinberger, Fontaine, and Simons (2011) who successfully provided evidence that inattentional blindness occurs in real-life situations. The main reason the current study did not yield similar results is because the current study focused on the influence of attentional sets in the possibility of noticing an unexpected event while the study by Chabris et al. merely tested the effects of attending to primary task on the failure to notice another event that occurs not too far away from an individual's visual field. Therefore, the difference in manipulation could have lead to the difference in results obtained.

The results of the present study is also not consistent with the study conducted by Lane (2006) who successfully found that performing two tasks simultaneously contributes to inattentional blindness. The present study failed to yield similar results as it took on a different approach in understanding the phenomena of inattentional blindness: the influence of attentional sets. On the other hand, the study by Lane focused on the effects of completing a single primary task versus completing two tasks: one primary and one secondary task, on the probability of noticing the unexpected stimulus. However, it can be argued that there are similarities in the two studies: both involved participants engaging in either one primary task (control group) or engaging in one primary task and one secondary task (primary: counting a stimulus, secondary: watching the video). Nevertheless, the failure in obtaining similar results could then be attributed to the fact that in the present study, both tasks were visual tasks while in the study by Lane, one task was a visual task while the other was an auditory task.

Lastly, the results obtained by Rivardo et al. (2011) did not match the current study findings. Rivardo et al. successfully found that consistent attentional sets positively affects an individual's ability to notice an unexpected stimulus. Although the current study replicated part of the methods from Rivardo et al., similar findings was still not obtained. This could be attributed to a few reasons. Firstly, unlike the current study, Rivardo et al. included a filler task: given a word list to memorise. This would have provided participants with a low level of expectancy while watching the video. However, this has its drawbacks as while watching the video, some participants may be preoccupied with rehearsing the word list silently instead of keeping tallies of the attentional sets they were assigned to. Besides that, the results obtained could also be attributed to the fact that all participants were exposed to the same video of a theft of a shopping bag. In the current study, in order to manipulate for the absence and presence of weapon, participants were exposed to either one of two videos. Therefore, as participants were exposed to two different videos, the effect of inattentional blindness could not have been accurately compared.

Although not all the hypothesis were supported, the present study still holds several strengths. For example, the questionnaire contained both multiple choice questions as well as open ended questions. Research has found that under free recall, individuals experience hypermnesia: exceptionally vivid recollection or memory (Dunning & Stern, 1992). The same researchers also found that providing multiple choice questions reduced hypermnesia (Dunning & Stern). Besides that, there was a control for a possible conflict of interest that could have arise. Specifically, participants were provided with an additional question asking them if they knew any of the actors in the video. If they answered yes, their data would have been discarded. If this was not done, the validity of the result obtained would have been compromised as participants' attention may have been fixated on the actors they knew. In addition to that, there was a condition of low expectancy among the participants as participants were told they were participating in an experiment studying the effects of task difficulty on performance. Therefore, there was a control for demand characteristic as the method of manipulation of the independent variables were implicit: they were not aware that the primary task they were attending to was the independent variable. Moreover, the same victim and perpetrator was used in both the videos.

However, just like any other research, the present study holds several shortcomings. One of the weaknesses of this study was the length of the videos. In the study by Rivardo et al. (2011), the videos shown to the participants were only three minutes long. The current study employed a seven minutes long video. The length of the video is believed to be a weakness as it might have caused boredom among participants as there were moments in which some participants were not paying full attention to the video after. This was observed by the researcher during the duration of the experiments. On another perspective, Kassin, Tubb, Hosch, and Memon (2001, as cited in Fahsing & Granhag, 2004) studied the effects of crime duration on eyewitness memory and found that the longer subjects observed an event, the more information is likely to be encoded and remembered. However, in the present study, the crime duration was barely 10 seconds despite the video being seven minutes long. Therefore, as according to the Kassin, Tubb, Hosch, and Memon (2001, as cited in Fahsing & Granhag, 2004), less information of the eyewitness event was able to encoded and recalled.

Besides the length of the video, another weakness that is evident in the current study is choreography of the videos. In essence, there are two weaknesses that relates to the choreography. Firstly, as the videos were filmed in a public restaurant, despite having research assistants acting out as according to the task they were assigned to, there were other patrons in the restaurants that were either entering or leaving the place. This weakness could not have been controlled for by the researcher. Therefore, both the videos were not similar in regards to the actions of the other patrons in the restaurant. The dissimilarity could have affected the validity of the results obtained as the perceptual load (the number of bags carried by individuals or the number of individuals in white coloured tops) may have been unequal in both the videos. Besides that, another weakness would be the choreography of the video focused participants' attention to the entrance of the restaurant as each actor came in one after another from the same direction. Moreover, the theft occurred at the entrance of the restaurant. To further illustrate this belief, Most, Simons, Scholl, & Chabris (2000) found significant results that the location of the unexpected stimulus affects the probability of noticing it. In particular, it has been found that observers are more likely to notice the unexpected stimulus if its near the focus of the individual's attention. Hence, as the focus of the individual's attention is predicted to be fixated at the entrance of the restaurant, this could have influenced the likelihood of participants noticing the unexpected event: theft.

Another flaw that can be found in the present study was the presence of an immediate recall task or the absence of a filler task. Experts in the area of eyewitness memory agree that the accuracy of memory decreases over time (Kassin et al., 2001, as cited in Fahsing & Granhag, 2004). Therefore, as participants' memory of the event was tested immediately after the video, it reduces the generalization of the results obtained as in an actual eyewitness event, eyewitnesses would not be asked to provide a statement immediately. Thus, a filler task could have increased the validity of the result obtained as participants' memory of the eyewitness event would not be as fresh as it is after watching the video.

Additionally, another flaw that is evident is the present study is the failure to include a response confidence rating for the memory questionnaire. Although it was previously argued that a multiple choice question reduces hypermnesia, such questions may have forced some participants to guess an answer which would in turn affect the validity of the results obtained. Hence, including a confidence rating would have provided the researcher with a better perspective of participants' answer which will ultimately improve the validity of the results obtained.

Furthermore, as Easterbrook's cue utilization hypothesis (1959) proposed that weapons illicit an increase in arousal levels in eyewitnesses which then fixates their attention to the weapon, it would have been wise to measure participants' arousal level before and after participants viewed the video. Doing so would increase the validity of the result obtained: the decrease in accuracy of the eyewitness event was due to the presence of weapon. However, some researchers believe that it not all weapons illicit higher arousal levels. For example, Kramer et al. (1990, as cited in?) conducted a series of experiments which provided conflicting results. In one experiment, it was reported that weapons in a scene illicit higher arousal levels whereas in another experiment although no significant higher arousal levels were found, the weapon focus effect was still present. Kramer et al. attributed this finding to that the weapon focus effect may have occurred because the weapons were unusual items in the scenes. Therefore, although the present study is more interested in the practical implications of the effects of weapon on eyewitness memory instead of the theoretical reason behind the effect, it would have been interesting to measure the arousal levels in order to verify the weapon focus effect.

Besides the shortcomings mentioned above, there are also several confounding variables that should be addressed. For example, gender has been found to play a role in the weapon focus effect. Shaw & Skolnick (2006) found that there was an own-gender bias in identifying a perpetrator of an eyewitness event when there was an absence of weapon. However, there appeared to be an opposite effect when there was a presence of weapon. Therefore, although significant results were obtained, one should err on the side of caution when generalizing the results.

In addition to gender differences, individual differences in working memory capacity has also been found to play a role in both weapon focus effect and inattentional blindness. Research on working memory capacity has found that individuals with a greater working memory capacity is able to store more detailed mental image of the original scenario (Jaschinski & Wentura, 2002). Meanwhile, past research that studied the effects of working memory capacity on inattentional blindness has found that individuals with a lower working memory capacity have a higher tendency to not process the unexpected stimulus. Therefore, they are unlikely to notice the unexpected stimulus which in turns affects their memory of it.

Age has also been found to be a confounding variable in inattentional blindness studies. A recent study by Graham & Burke (2011) found that older adults are more likely to experience inattentional blindness as compared to younger adults (16-22). This could be attributed to a reduced attentional capacity in older adults which does not allow for a more demanding dual attentional tasks. Therefore, once again it is best to err on the side of caution when generalizing the results found to the rest of the population.

Meanwhile, individual's mood has also been found to influence one's attention. Specifically, a study by Becker & Leinenger (2011) found that individuals are more likely to notice an unexpected stimulus if it matches the observer's current mood. For example, in the current study, if participants were already anxious, they would be more likely to notice the theft as the victim displayed anxiety upon the theft. The findings from Becker & Leinenger suggest that one's mood can influence their attentional filter, thus affecting the information one pays attention to.

Besides confounding variables, there also appears to be other variables that should be taken into consideration if the results from the study is to be generalized or replicated. For example, it has been found that the type of weapon in a crime influences the strength of the weapon focus effect (Fahsing & Granhag, 2004). Specifically, it has been found that guns produce higher threat and arousal levels as opposed to knives. Thus, individuals exposed to firearms would provide poorer description of the perpetrator as compared to participants exposed to knives. However, opposite results have also be found. In an experiment by Fahsing & Granhag, they found that exposure to guns were associated with a higher accuracy in regards to description of the perpetrator as compared to exposure to knives. Despite the conflicting results obtained, one idea that is certain is that the type of weapons in crimes have an effect on the accuracy of eyewitness memory. Therefore, the results obtained in regards to the weapon focus effect should be read with caution as its effects may not be generalized to every eyewitness event.

In addition to the type of weapon, the number of perpetrators has also been found to influence eyewitnesses' memory. For example, Sporer (1996, as cited in Fahsing & Granhag, 2004) found that crimes with a higher number of perpetrators results in a lower accuracy of eyewitness memory due to divided attention: attention is divided among all the perpetrators. Furthermore, witness role has also been found to affect eyewitness memory. For instance, Christianson and Hubinette (1993, as cited Fahsing & Granhag, 2004) revealed that victims of crimes were better at recalling the eyewitness event as opposed to bystanders. As such, participants in the present study acted as bystanders, thus, their memory of the event is believed to be poorer.

Therefore, based on the shortcomings and confounding variables in the present study, a few improvements or extensions can be made for future studies. Firstly, the length of the video should not be too long in order to ensure that participants are fully attentive to the video. Besides that, the choreography of the video should be planned in a way that ensures participants focal point of attention is not directed at one location. This can be done by ensuring actors are walking into the scene and exciting the scene from various locations. Additionally, in order to really test witnesses' memory, it is advised to include a filler task or if possible, test their memory after a week. This will ensure the data generated can be applied to real-life situations. Furthermore, response confidence ratings should also be included in order to increase the validity of results obtained. Meanwhile, in order to control for confounding variables such as gender, individual differences in working memory capacity, age, and mood, appropriate measures need to be taken in order to fully validate the effects of the independent variables on the dependent variable. For example, participants should be provided with a mood scale before and after the experiment in order to study if their moods have an effect in the results obtained. On the other hand, variables such as type of weapon and number of perpetrators should be studied together with weapon focus and inattentional blindness in order to see if a difference in results exists.

As mentioned in the beginning of this research paper, there are several implications of conducting this study. Firstly, by studying the effects of both weapon focus and inattentional blindness on eyewitness memory can significantly reduce the number of unreliable eyewitnesses. This is possible as the current study aimed to create a simulated eyewitness event whereby participants acted as bystanders who were busy with a task at hand instead of waiting for a cr