Integrating inattentional blindness and eyewitness memory.
Inattentional blindness (IB) and eyewitness memory are two extensive areas of research, yet there has been little integration of these related areas. The purpose of this study was to examine how attentional set can affect IB for a simulated eyewitness event and subsequent memory for it. 187 undergraduates watched a video of the food court at a local mall. During the video, a shopping bag was unexpectedly stolen. Participants who were given the task of counting mall patrons with shopping bags (consistent attentional set) were more likely to notice the theft (38%) than participants counting shoppers wearing blue shirts (inconsistent attentional set, 19%) and participants without a counting task were most likely to notice the theft (90%). Participants who were inattentionally blind to the theft were less accurate in reporting details not mentioned in a narrative, accurately reported in a narrative, and misled in the narrative than those who noticed the theft (ps < .03). Laboratory studies using simulated eyewitness events without an additional distracting task may fail to account for attentional factors that can affect accuracy of eyewitness accounts.

Article Type:
Shopping malls (Cases)
Theft (Cases)
Eyewitness identification (Cases)
Rivardo, Mark G.
Brown, Kelly A.
Rodgers, Alexis D.
Maurer, Sara V.
Camaione, Tyler C.
Minjock, Robert M.
Gowen, Gina M.
Pub Date:
Name: North American Journal of Psychology Publisher: North American Journal of Psychology Audience: Academic Format: Magazine/Journal Subject: Education; Psychology and mental health Copyright: COPYRIGHT 2011 North American Journal of Psychology ISSN: 1527-7143
Date: Nov, 2011 Source Volume: 13 Source Issue: 3
Event Code: 980 Legal issues & crime Advertising Code: 94 Legal/Government Regulation Computer Subject: Company legal issue
Product Code: 9101323 Theft NAICS Code: 92212 Police Protection
Geographic Scope: United States Geographic Code: 1USA United States
Accession Number:
Full Text:
Although there is abundant literature on both inattentional blindness (IB) and eyewitness memory as separate areas of research, there has been little focus on integrating these two. Laney and Loftus (2010) made a strong case for conducting research that combines IB and change blindness (CB) with eyewitness memory. In eyewitness studies that are conducted in laboratory settings, the participants are usually expecting to see something happen and are aware that they will be questioned about the material. In approximately two thirds of eyewitness studies reviewed, researchers showed participants a video, slideshow or series of photos and implied or explicitly told them they would be asked questions about the stimuli. These instructions may have led them to pay closer attention to the events portrayed in the materials than they would have if they had been actual witnesses to the real-life event. This heightened level of expectancy in many of the studies may have resulted in greater frequencies of noticing than would be expected outside of the laboratory where real witnesses are often engaged in some other task or activity when the event occurs. Others (Ihlebaek, Love, Eilertsen, & Magnussen, 2003; Malpass, Sporer, & Koehnken, 1996) have acknowledged these problems with traditional eyewitness studies and Lane (2006) directly addressed effects of concurrent tasks on eyewitness memory. She found that participants who engaged in a secondary task were more susceptible to misinformation presented after viewing a slide show. However, her secondary task was auditory, unrelated to the eyewitness slide sequence, and did not produce inattentional blindness.

IB is the failure to notice an unexpected stimulus because attention is focused on another task or object (Mack & Rock, 1998). A few studies have directly addressed how IB and the related phenomenon of CB are relevant in criminal cases. Chabris and Simons (2010) suggested Boston police officer Kenny Conley may have been inattentionally blind to the beating of African-American undercover officer Michael Cox by uniformed officers who mistook Cox for a suspect. Conley, who reportedly ran by within feet of the beating, claimed to have not noticed the assault. Chabris, Weinberger, Fontaine and Simons (2011) simulated this situation in a recent study. They found that 35% of undergraduates following a fellow jogger at night failed to notice three students engaged in a staged fight. Davis, Loftus, Vanous, and Cucciare (2007) conducted a study on CB, which occurs when a change to a visual scene goes undetected. Participants watched a video of a theft in a grocery store. As in typical IB studies, participants were given a task to complete (memorizing items from aisles) while watching the video. Sixty-four percent of participants failed to notice that the person who emerged from behind a display was not the thief who had recently gone behind it, and overall misidentifications were over 70%. The visual attention literature, encompassing both IB and CB, addresses issues pertinent to eyewitness memory performance, such as failure to encode features of a scene or failure to distinguish differences properly. Incorporating aspects of the IB paradigm, like dual task procedures, may result in findings that are more applicable to real-life eyewitness situations where bystanders are not maximally attentive to the important event occurring in front of them.

Since Mack and Rock (1998) first coined the term, there have been numerous studies to demonstrate the applicability of IB to various situations. In a well-known study building on Neisser (1979), Simons and Chabris (1999) instructed participants to count basketball passes between team members wearing either white or black shirts in a video. While players passed the ball, a woman in a gorilla costume or a tall woman with an umbrella appeared on screen for 5 seconds, walking through the middle of the screen. Despite the salience of the event, only 54% of participants noticed.

A variety of tasks and stimuli have been used in IB studies, such as following shapes on a computer screen (Most et al., 2001; Most, Scholl, Clifford, & Simons, 2005) or making decisions during a sports simulation (Memmert & Furley, 2007). Others have explored more serious consequences. Participants in driving simulators have missed signs or pedestrians because they were talking on a cell phone (Strayer, Cooper & Drews, 2004; Strayer, Drews & Johnston, 2003) and have been slow to notice an oncoming motorcycle when attending to traffic signals (Most & Astur, 2007), and pilots in a flight simulator have failed to see a plane on the runway they were landing on (Haines, 1997).

Most et al. (2005) expanded on the perceptual cycle framework (Neisser, 1979) that describes how instances of attention capture and awareness capture can occur at different stages of processing. Attention capture is the shifting of unconscious attention to stimuli due to their salience, and does not require the capture of awareness. Awareness capture refers to when stimuli are acknowledged on the conscious level (Most et al., 2005; Neisser, 1979). An individual's attentional set consists of the specific characteristics of stimuli to which an individual is devoting his or her attention at any given time. In IB research, the attentional set is established by the instructions given for the primary task. The occurrence of IB is affected by attentional sets based on color (Most & Astur, 2007; Most et al., 2001; Simons & Chabris, 1999), shape, luminescence, motion (Most et al., 2001), number (White, & Davies, 2008), and semantic meaning (Koivisto & Revonsuo, 2007). Whether a stimulus captures awareness depends on the interaction of salience and attentional set. Some stimuli may capture attention, but unless they are consistent with the attentional set, they will not capture awareness. Other stimuli may be so salient that they capture awareness regardless of the attentional set (Most et al., 2005).

Karns and Rivardo (2010) found attentional set for expected action could affect noticing of an unexpected fight between two students. Participants who watched surveillance video in an attempt to find a person known to have been in physical confrontations with another student were more likely to notice a physical confrontation between two students than to notice a conspicuous person in a gorilla suit. Participants who were told to look for the target person to inform him about a family emergency were more likely to notice the gorilla than the confrontation. In each of these studies, only the effect of attentional set on noticing the unexpected event was examined; the accuracy of participants' memory for the features of the unexpected object or the details of the unexpected event were not examined.

Clearly, the IB literature shows that salient stimuli can pass through one's visual field without being noticed and these findings are relevant to eyewitness memory. If witnesses are not able to notice something that occurs right in front of them, then they are incapable of correctly reporting the incident. In real life situations, people are likely to be engaged in other tasks when eyewitness events occur so IB may lead to worse memory performance than has been reported in the eyewitness literature.

Eyewitness memory is often inaccurate and impressionable. In laboratory studies, researchers can affect accuracy of eyewitness memory by using techniques that result in the misinformation effect. The misinformation effect refers to changes in a person's memory for past events due to exposure to incorrect information (for review see Loftus, 2005). Misinformation can lead to alterations in the reporting of details of an event or to the creation of completely false memories for events that never occurred.

Through the lost in the mall technique (Lindsay, Hagan, Read, Wade & Garry, 2004; Loftus, 1993; Loftus & Pickrell, 1995), researchers have created false memories for events that never occurred. Examples of such false memories include childhood memories of being lost in a mall as a child (Loftus & Pickrell), being hospitalized overnight, spilling punch on the parents of a bride at a wedding, evacuating a grocery store because the overhead sprinklers were mistakenly set off, causing a car accident in a parking lot by releasing the parking brake on a car (Hyman, Husband, & Billings, 1995) and riding in a hot air balloon (Wade, Garry, Read, & Lindsay, 2002). Using a mock advertisement, Braun and colleagues led participants to believe that they had met Bugs Bunny at Disney World even though Bugs Bunny is a Warner Bros character and would not be at Disney World (Braun, Ellis, & Loftus, 2002; Braun-LaTour, LaTour, Pickrell & Loftus, 2004). Given the ease with which memories of events can be created, it is easy to see how participants who are initially inattentionally blind to the event might form memories for that event based upon information provided later.

Misleading post event information can be presented in a variety of ways. Post event narratives containing misinformation can lead to the false reporting of misled details (e.g. Gabbert, Memon, Allan, & Wright, 2004; Searcy, Bartlett, & Memon, 2000; Wright & Stroud, 1998). Leading questions can also cue distortions in several ways. The intensity of verbs used in the question can affect speed estimates (Loftus & Palmer, 1974) and misinformation can be implied in questions, for example by asking if participants saw children getting off the bus when a bus was never shown (Loftus, 1975; Loftus, Miller, & Burns, 1978). Others have found misinformation effects through co-witness collaboration (e.g. Gabbert, Memon, & Allan, 2003; Luus & Wells, 1994).

Although Davis et al. (2007) examined change blindness in an eyewitness memory study, no known published study has been found that has combined the techniques of IB and eyewitness memory research. In typical eyewitness studies, participants who do not notice the critical event are removed from the data pool. We included the individuals who were inattentionally blind and we attempted to create false memories of the critical event that they did not notice by presenting false information in a post event summary of the video. Participants who did notice the event were also exposed to the misinformation and were expected to show misinformation effects as well.

As in all IB experiments, participants were given a primary, attentional task to complete and were exposed to an unexpected event. Participants watched a video portraying a food court at a local mall. The primary task varied in the attentional set that it induced in participants. The unexpected event was the theft of a young woman's shopping bag by a young man. Participants in the consistent attentional set condition were told to count shopping bags, those in the low consistency attentional set condition were told to count shoppers wearing blue shirts, and those who were given no instructions to count represented the control group, which was meant to represent the type of attentional set typically induced in eyewitness memory studies.


Participants were also given instructions that would manipulate their level of expectancy. Some participants were told they would be asked about content of video after they watched it (high expectancy) and some were just told to watch it while they waited to recall a word list (low expectancy). The high expectancy control group was most similar to what participants in traditional eyewitness studies experience; they knew that they would be asked about the events in the video and they were not given a primary attentional task. After answering questions to assess whether they were inattentionally blind, participants read a brief summary of the theft that contained some true and some false information before answering additional questions about the video.

We hypothesized that IB and misinformation effects would coincide--conditions that produced higher rates of IB should have also produced larger misinformation effects and lower accuracy. We expected participants in the high expectancy condition to notice the theft more often than those in the low expectancy condition because they were expecting to be asked about the video. Those in the high expectancy condition should also have a smaller misinformation effect because their instructions indicated they would be asked questions after watching the video, which should have caused them to pay more attention to the details, increasing the chances that they would have a memory trace for the theft. As for the attentional set manipulation, we hypothesized that participants in the consistent attentional set condition would notice the theft more often than participants in the inconsistent condition. According to the attentional set theory, participants in the consistent condition should have noticed the theft more because their task of counting shopping bags created sensitivity for noticing shopping bags, which matched the critical event, the theft of a shopping bag. Also, the participants in the consistent condition should have had a smaller misinformation effect than those in the inconsistent condition because they were most likely to notice the theft, which is what the misinformation concerned. The participants in the control condition should have scored between the consistent and inconsistent groups for dependent measures because they were not influenced to pay more or less attention to the critical event. However, being free from counting might make them more likely to detect the theft than if they were counting shirts, and it might also make them more likely to notice the theft than if they counted shopping bags simply because they did not need to devote any attention to another task and were expecting something interesting to occur. As for the interaction of the attentional set and expectancy, the aforementioned pattern for attentional set was expected to hold true for the consistent and inconsistent conditions but the control conditions were expected to differ from each other based on expectancy. Participants in the control condition did not count anything, so the expectancy manipulation was expected to have a greater effect. Participants in the high expectancy control condition were predicted to notice the theft more often and be much less susceptible to the misinformation effect than those in the low expectancy control condition.



One hundred eighty-seven undergraduate students (131 female) from Saint Vincent College participated individually in groups of up to 12. Students were recruited from classes across disciplines. They were randomly assigned to 1 of 12 combinations of conditions. Some participants received extra credit for their participation.


The present study incorporated a 3 (attentional set) x 2 (expectancy) x 2 (question type) factorial design for mixed groups. The attentional set manipulation relates to the similarity between what participants were told to count and the critical item, the shopping bag involved in the unexpected theft. In the consistent attentional set group, participants were asked to count the number of individuals in the video, keeping separate counts for those carrying shopping bags and those without shopping bags. In the inconsistent attentional set group, participants were asked to count the number of shoppers wearing blue shirts and not wearing blue shirts in the video. In the control condition, participants were not told to count anything. The expectancy manipulation concerned whether participants were told that they were going to be asked general questions about the content of the video. In the high expectancy condition, participants were told that they would be asked questions about the video's content after they viewed it, but in the low expectancy condition participants were not told that they would be asked about the content of the video, they were only told to watch it. The question type refers to the misinformation that was presented to participants in the short narrative. All participants received misinformation about two of seven details of the theft but to control for the variable, two different forms were used.

Materials & Procedures

After providing their written consent to participate, participants received a packet. The first page in the packet was the word list and participants were given 1 min to study and memorize the list. The word list contained 20 words (see Appendix), which participants were told to memorize. The words were checked against free association norms to verify that they were not related (Nelson, McEvoy, & Schreiber, 2004). Participants were told that they would be given a free recall test on the words after they watched a video. The free recall served several purposes. First, it served to delay the recall of information from the video. Secondly, it gave credibility to the low expectancy control condition because the instructions to participants in that condition implied that the purpose of the video was to delay the free recall of the word list, thereby giving participants a reason to watch the video without giving them reason to believe they would be tested on the content. Participants then read the instructions pertaining to the video they were about to watch according to the expectancy and attentional set conditions to which they were randomly assigned. Participants in the consistent attentional set condition received instructions that directed them to count the number of individuals carrying shopping bags and the number who were not, and participants in the inconsistent attentional set condition were told to count the number of individuals wearing and not wearing blue shirts. Participants who were instructed to count were told to use tally marks to keep track. Because the number of shopping bags and blue shirts were not equal and high working memory load (Fougnie & Marois, 2007) and perceptual load (Macdonald & Lavie, 2008; White & Davies, 2008) have been shown to increase IB (but see Koivisto & Revonsuo, 2009) , participants were asked to count individuals who both met and did not meet the criterion so that the attentional load would be consistent across conditions.

After reading their instructions and being permitted to ask questions, participants watched a 2 min and 34 s video that was recorded in the food court of a local shopping mall. In the video a college age man (thief) in a white t-shirt sat reading a newspaper in the foreground with shoppers walking by or buying food from a pizza shop in the background. At 24s a young woman (victim) placed her shopping bag in a chair at a table adjacent to the one the young man was using and then sat down across the table from it. She texted until 1:05 and then casually watched the people walk by and occasionally checked her cellular telephone. At 2:13 a friend walking buy noticed her, at which point she stood up and they hugged. At 2:19 the thief glanced at the women as they stood talking with their backs to the shopping bag. He then stood up and walked past the victim's table casually grabbing the shopping bag at 2:25 and disappearing off screen to the left at 2:27. At 2:30, the victim and friend sat down at the table with the friend sitting in the chair where the bag had been left. At 2:33 the victim expressed surprise that her shopping back was missing before the video ended with a grey static screen.

After watching the video, participants completed the rest of the packet at their own pace. Participants in the consistent and inconsistent attentional set conditions were asked to total their tallies. After writing as many of the 20 words from the word list as they could recall, participants responded to 7 inattentional blindness based questions about the content of the video. The first question asked participants to describe, in their own words, what happened in the video. The questions became increasingly specific, with the last question asking participants if they noticed the theft of a shopping bag. Then participants read a two-sentence narrative of the theft. We selected four details to mislead participants about, but each participant was only misled on two of the four. The accurate details are presented in parentheses in the narrative below.

Following the narrative were 7 detailed questions about the theft. Participants were then asked to fill out a demographic questionnaire. Lastly, they were given an 8-item Objective Eyewitness Questionnaire (Appendix) that asked additional questions about the theft, including questions about the misinformation, and asked participants to rate their confidence in the accuracy of their answers using a 4-point Likert scale.


Several analyses were conducted to assess the hypotheses. First, to see whether instances of IB were more likely to occur under particular attentional set and expectancy conditions, Chi-square analyses were conducted. Next, ANOVAs were conducted to determine whether recall accuracy and susceptibility to misinformation were related to whether participants noticed the theft. Third, an ANOVA was conducted to determine whether recall accuracy was affected by attentional set, expectancy, and information type (misled, not misled, not mentioned), which essentially enabled us to determine if the misinformation effect varied by attentional set and expectancy. Next, effects on confidence and accuracy or recall were examined and finally an ANOVA was conducted to determine whether confidence varied as a function of attentional set and being inattentionally blind.

IB was scored by counting participants who claimed to have seen the theft as noticing the event and those who claimed not to see the event as being inattentionally blind. Chi-square analyses were conducted to determine if noticing differed by conditions (see Table 1). Frequency of IB in the low expectancy, [chi square] (2, 93) = 28.38, p < .001 and high expectancy, [chi square] (2, 94) = 41.16, p < .001, conditions varied by attentional set. Fisher's Exact tests revealed that in both the high and low expectancy conditions, participants were inattentionally blind less often in the control condition than in either the inconsistent or consistent attentional set conditions (all p's < .001). Differences in IB between the consistent and inconsistent attentional set conditions were not significant for the low (p = .10) and high (p = .064) expectancy conditions. No differences in IB based on expectancy were found (all p's > .33), so data were collapsed across expectancy conditions. Subsequently, the consistent attentional set condition was found to have a lower frequency IB than the inconsistent attentional set condition (p = .015).


How inattentional blindness was related to eyewitness accuracy (proportion correct) and misinformation reporting was evaluated with a factorial ANOVA for mixed groups and independent samples t-tests. Overall, accuracy varied as a function of information type, F(2, 340) = 80.35, p < .001, [[eta].sub.p.sup.2] = .32. Accuracy for misled information (M = .81, SD = .25) was lower than for both information not misled (M = .94, SD = .17) and information not mentioned (M = .88, SD = .21; both ps < .001), and accuracy for information not mentioned was lower than accuracy for not misled information (p = .001). The interaction of information type and IB was also significant, F(2, 340) = 16.38, p < .001, [[eta].sub.p.sup.2] = .09 As shown in Figure 2, participants who were inattentionally blind were especially susceptible to the misinformation (p < .001) but differences for information not mentioned in the narrative (p < .001) and not misled (p = .023) were also reliable. When making an error on questions pertaining to misled information, sometimes participants gave novel responses rather than reporting the misled information. An independent samples t-test revealed that participants who were inattentionally blind reported the specific misinformation 45% (SD = 38.8) of the time compared to participants who noticed the theft who reported the misinformation just 8% (SD = 18.8) of the time, t(130.84) = 8.09, p <.001.

Participants who were inattentionally blind to the theft were less accurate and reported more misinformation. A mixed factorial ANOVA was conducted to examine differences in recall accuracy (proportion correct) by information type (misled, not misled, not mentioned in the narrative), attentional set (control, inconsistent, consistent), and expectancy (high, low). Information type affected recall accuracy, F(2, 332) = 79.78, p < .001, [[eta].sub.p.sup.2] = .32. Responses pertaining to misled information were less accurate (M = .64, SD = .38) than not misled (M = .94, SD = .17) and not mentioned (M = .88, SD = .21; both p's < .001), which also differed from each other (p = .001). Expectancy did not affect recall accuracy, F(1, 166) = 0.04, p = .84. Recall accuracy of participants in the high expectancy (M = .82, SE = .02) and low expectancy (M = .81, SE = .02) conditions did not differ. Recall accuracy varied by attentional set, F(2, 166) = 16.12, p < .001, [[eta].sub.p.sup.2] = .16. Accuracy in the control condition (M = .91, SE = .02) was greater than in the inconsistent (M = .75, SE = .02) and consistent (M = .78, SE = .02) conditions (both p's < .001), but the consistent and inconsistent conditions did not differ from each other (p = .45).

The only interaction that reached significance was that of similarity and information type, F(4, 332) = 4.00, p = .004, [[eta].sub.p.sup.2] = .05. As reported above, accuracy for misled information was lower than for information that was not misled or not mentioned in the narrative. As shown in Figure 3, the difference in accuracy between misled information and the other two types of information was greater in the inconsistent and consistent conditions than in the control condition. The information type by expectancy, F(2, 332) = 1.38, p = .25, similarity by expectancy, F(2, 166) = 0.22, p = .80, and information type by attentional set by expectancy, F(4, 332) = 0.56, p = .69, interactions failed to reach significance.

The previous analysis examined accuracy for misled information, but errors included cases where participants reported misinformation (e.g. victim ate a sandwich) and cases where a novel response was given (e.g. the victim read a magazine). Another ANOVA was conducted on the proportion of misinformation that was explicitly reported by participants. Expectancy did not affect misinformation acceptance, 174) = .932, p = .34, but similarity did, F(2,178) = 10.65, p < .001, [[eta].sub.p.sup.2] = .11.


Participants in the control condition (M = .11, SD = .21) were less likely to report misinformation than those in the consistent (M = .35, SD = .38) to reach significance, F(2, 174) = .05, p = .95.

Accuracy for correct responses was coded 1 and accuracy for incorrect responses was coded -1. This value was multiplied by the participant's reported confidence in the particular response. Values close to 4 represent highly confident and correct and values close to -4 represent highly confident but wrong. Scores closer to zero indicate lower confidence. As expected, confidence x accuracy differed by information type, F(2, 330) = 67.60, p < .001, [[eta].sub.p.sup.2] = .29. Responses pertaining to misled information (M = 1.16, SD = 2.46) were lower than for both information not mentioned (M = 2.28, SD = 1.63) and non-misled information (M = 2.99, SD = 1.33), which also differed from each other (all ps < .001). Scores also differed by attentional set, F(2, 165) = 19.39, p < .001, [[eta].sub.p.sup.2] = 19. Reflecting the same pattern found in previous analyses, scores in the control condition (M = 2.94, SE = 0.17) were lower than in both the inconsistent (M = 1.49, SE = 0.18) and consistent (M = 1.90, SD = 0.17) conditions (both ps < .001). However, the consistent and inconsistent conditions did not differ from each other (p = .096). As in previous analyses, the effect of expectancy was not significant, 165) = 0.09, p = .76. The low (M = 2.08, SE = 0.14) and high (M = 2.14, SE = 0.14) expectancy conditions did not differ in accuracy x confidence. Only the interaction of information type and attentional set (see Figure 4) was significant, F(4, 330) = 3.26, p = .012, [[eta].sub.p.sup.2] = .04 (all other ps > .20). For each condition of attentional set all differences between information types were significant at the .01 level except in the control condition where the difference between not mentioned and not misled was less reliable (p = .045). Across attentional set conditions the control condition was higher than the other two conditions for misled and not mentioned (all ps < .001). Not misled scores did not differ between the control and inconsistent (p = .058) and control and consistent (p = .432) conditions. This pattern is similar to that found with overall accuracy except that accuracy x confidence was lower in the two counting conditions than the control condition, whereas for overall accuracy those differences were not found.


Finally, an ANOVA was conducted to determine if participants who were inattentionally blind were less confident in their responses and to see how this affect may have varied by attentional set. Participants who were inattenionally blind (M = 2.62. SD = 0.68) were less confident than those who noticed the theft (M = 3.49, SD = 0.65), 180) = 49.25, p < .001, [[eta].sub.p.sup.2] = .22. Neither attentional set, F(2, 180) = 0.64, p = .53, nor the interaction of IB and attentional set, F(2, 180) = 1.58, p = .21, were significant.


Fifty-one percent of the participants failed to notice the theft. On the surface this rate appears to match rates found in the IB literature (e.g. Simons & Chabris, 1999), however in our experiment roughly one third of the participants had no secondary task to draw their attention away from the unexpected event. Seventy-two percent of the participants who did engage in a counting task were inattentionally blind; our unexpected event may have been more difficult to notice than those used in previous research.

Unexpectedly, the frequency of inattentional blindness did not vary by expectancy condition. We predicted participants who were not told that they would be asked questions about the content of the video beyond those pertaining to the counting task would be less likely to notice the theft. However, during debriefing many participants reported that they did expect something unusual to happen regardless of which condition they were in, partially because other participants in the session received instructions to tally their counts during the video. It is difficult to convince suspicious research participants that they are watching a video that was obviously recorded for the experiment, rather than a clip from a television show for example, just to pass the time. Fortunately, we used another technique to better approximate the attentional conditions that might be experienced by actual eyewitnesses; we gave approximately two thirds of the participants a counting task to occupy some of their attention.

As expected, IB did vary by attentional set. Having a counting task reduced participants' ability to notice the theft, and when data were collapsed across expectancy conditions the difference between the consistent and inconsistent attentional set conditions was reliable; participants who were watching to see whether mall visitors were carrying shopping bags were more likely to notice the theft of a shopping bag than participants who were attending to the color of shirts shoppers were wearing. These results are consistent with prior findings concerning the effects of attentional set on IB (e.g. Most & Astur, 2007; Most et al., 2001; Simons & Chabris, 1999) and more specifically those concerning the effects of attentional sets based on more than simple physical characteristics (Karns & Rivardo, 2009; Koivisto & Revonsuo, 2007). Note that there was no particular reason for participants to be attending to the bag as it was stolen because it had appeared in the video, and likely been counted, a full 2 min prior to the theft.

These findings have several implications for eyewitness memory research. First, IB may reduce the number of reliable witnesses. We argue that in real eyewitness situations witnesses are likely to be doing or thinking about something else when the event occurs; they are unlikely to be just watching the world around them waiting for a car accident, a crime, or someone in a gorilla suit to walk by. In addition, given high expectations that "something will happen" when participants are asked to watch a video, a secondary task of some sort is needed to better simulate an eyewitness situation. When a secondary task is employed, we see that inattentional blindness may reduce the number of reliable witnesses considerably. Furthermore, participants who were inattentionally blind to the theft were less accurate on all types of questions about it, not just those for which misinformation was presented.

One might argue that in real eyewitness situations bystanders who did not see the event are not considered to be witnesses by law enforcement personnel and in the present study those who were IB were less confident in their responses. However, cases where witnesses initially claim not to have seen anything, or not to have been in a position to see the event very well and then later testify with high confidence, have been documented for decades (e.g. Loftus, 1979).

The manipulation of attentional set also affected accuracy and misinformation reporting directly, but these effects were limited to information that was misled in the narrative. Accuracy for the misled information was lower in the consistent and inconsistent attentional set conditions than in the control condition and the explicit reporting of that misled information was greater in the consistent and inconsistent conditions. Having a counting task reduced performance on these measures but varying the focus of the counting task did not have an appreciable effect. Although the effects of attentional set were strong enough to reveal differences in IB, they did not translate into overall differences in susceptibility to misinformation.

The limited effects of attentional set on accuracy and the high level of accuracy for inattentionally blind participants were surprising. Even participants who reported not noticing the theft, were correct 79% of the time on questions about the incident for which no information was presented in the narrative. By the end of the video, the thief may have been very familiar to participants. He was visible from the beginning of the video until he exited the scene with the shopping bag. Of all the individuals appearing in the video, the thief was present the longest and because he was in the foreground he occupied the most screen space. Even participants who did not notice the theft were made aware of it through the narrative. In previous research on unconscious transference (see review and meta-analyses by Deffenbacher, Bornstein, & Penrod, 2006) participants have identified an innocent person who was at the scene or seen previously as the suspect. Similarly, participants in this experiment may have described the thief merely because he was the most familiar person from the video. In trying to construct a memory for a theft they did not notice, they used the information they could readily recall. To test this hypothesis in a follow-up, an innocent person could be present for a large portion of the video with the thief making a brief appearance.

In the present study we attempted to simulate an eyewitness situation in a laboratory type study. In previous studies, researchers have seldom engaged participants in other tasks that would consume attentional resources as would be the case in a real eyewitness situation. As Laney and Loftus (2010) suggested, the IB literature is an appropriate place to look to learn more about how distractions might hinder eyewitness accuracy. We were successful in replicating effects of attentional set on IB in the eyewitness memory paradigm and found that participants who were inattentionally blind were more susceptible to misinformation. Despite high rates of IB, participants were very accurate in describing the thief, which on the surface appears to suggest IB may not have lasting effects on the accuracy of eyewitness memory. However because the thief was on screen in the foreground for nearly the entire video participants may have implicitly or explicitly identified him as the thief after they were told a theft had occurred. In a scenario where the thief is just another passerby, an innocent bystander may be identified as the perpetrator. Our findings serve to partially bridge the gap between the IB and eyewitness memory literature, but also call attention to the need to consider specifically whether IB may increase instances of unconscious transference.


Word List

Plate, Door, Green, Teach, Prime, Aardvark, Foot, Silent, Left, Box, Sticker, Chain, Mascara, Lay, Absurd, Pen, Demand, Table, Increase, Fast

Objective Eyewitness Questions

1. What color was the stolen shopping bag?

2. What color was the shirt of the shopping bag's owner?

3. What was the owner doing prior to being approached by her friend?

4. What color was the shirt the thief was wearing?

5. What was the thief doing before stealing the shopping bag?

6. What was the sex of the thief? Male Female

7. In which direction did the thief walk after stealing the shopping bag?

Toward camera Away from camera

Note: Robert Minjock is now at the Department of Psychology, Central Michigan University. Gina Gowen is now at the Counseling Department, Franciscan University of Steubenville.


Braun, K., Ellis, R., & Loftus, E. (2002). Make my memory: How advertising can change our memories of the past. Psychology & Marketing, 19(1), 1-23.

Braun-LaTour, K. A., LaTour, M. S., Pickrell, J. E., & Loftus, E. F. (2004). How and when advertising can influence memory for consumer experience. Journal of Advertising, 55(4), 7-25.

Chabris, C. F., Weinberger, A., Fontaine, M., & Simons, D. J. (2011).You do not talk about Fight Club if you do not notice Fight Club: Inattentional blindness for a simulated real-world assault. i-Perception, 2(2), 150-153. doi:10.1068/i0436.

Chabris, C.F., & Simons, D.J. (2010). The Invisible gorilla and other ways our intuitions deceive us. New York: Crown.

Davis, D., Loftus, E., Vanous, S., & Cucciare, M. (2008). 'Unconscious transference' can be an instance of 'change blindness.' Applied Cognitive Psychology, 22(5), 605-623.

Deffenbacher, K. A., Bornstein, B. H., & Penrod, S. D. (2006). Mugshot exposure effects: Retroactive interference, mugshot commitment, source confusion, and unconscious transference. Law and Human Behavior, 30, 287-307.

Fougnie, D., & Marois, R. (2007). Executive working memory load induces inattentional blindness. Psychonomic Bulletin & Review, 14(1), 142-147.

Gabbert, F., Memon, A., & Allan, K. (2003). Memory conformity: Can eyewitnesses influence each other's memories for an event? Applied Cognitive Psychology, 17(5), 533-543.

Gabbert, F., Memon, A., Allan, K., & Wright, D. B. (2004). Say it to my face: Examining the effects of socially encountered misinformation. Legal and Criminological Psychology, 9(2), 215-227. doi:10.1348/1355325041719428

Haines, R.F. (1997). A breakdown in simultaneous information processing. Presbyopia Research: From Molecular Biology to Visual Adaptation. Eds K. Akins, volume 2 ( New York: Oxford University Press) 89-110.

Hyman, I., Husband, T., & Billings, F. (1995). False memories of childhood experiences. Applied Cognitive Psychology, 9(3), 181-197.

Ihlebaek, C., Love, T., Eilertsen, D. E., & Magnussen, S. (2003). Memory for a staged criminal event witnessed live and on video. Memory, 11, 319-327.

Karns, T. E., & Rivardo, M. G. (2010). Noticing of an unexpected event is affected by attentional set for expected action. North American Journal of Psychology, 12(3), 637-650.

Koivisto, M., & Revonsuo, A. (2007). How meaning shapes seeing. Psychological Science, 18(10), 845-849.

Koivisto, M., & Revonsuo, A. (2009). The effects of perceptual load on semantic processing under inattention. Psychonomic Bulletin & Review, 16(5), 864-868. doi:10.3758/PBR.16.5.864

Lane, S. M. (2006). Dividing attention during a witnessed event increases eyewitness suggestibility. Applied Cognitive Psychology, 20(2), 199-212. doi:10.1002/acp.1177

Laney, C., & Loftus, E.F. (2010). Change blindness and eyewitness testimony. In G. M. Davies & D. B. Wright (Eds.), New frontiers in applied memory. Psychology Press, p. 142-159.

Lindsay, D., Hagen, L., Read, J., Wade, K., & Garry, M. (2004). True Photographs and False Memories. Psychological Science, 15(3), 149-154.

Loftus, E. F. (1975). Leading questions and the eyewitness report. Cognitive Psychology, 7(4), 560-572.

Loftus, E. F. (1979). Eyewitness Testimony. Cambridge, MA: Harvard University Press.

Loftus, E.F. (1993). The reality of repressed memories. American Psychologist. 48, 518-537.

Loftus, E. F. (2005). Planting misinformation in the human mind: A 30-year investigation of the malleability of memory. Learning & Memory, 12(4), 361-366.

Loftus, E., Miller, D., & Burns, H. (1978). Semantic integration of verbal information into a visual memory. Journal of Experimental Psychology: Human Learning and Memory, 4(1), 19-31.

Loftus, E. F., & Palmer, J. C. (1974). Reconstruction of automobile destruction: An example of the interaction between language and memory. Journal of Verbal Learning & Verbal Behavior, 15(5), 585-589. doi:10.1016/S00225371(74)80011-3

Loftus, E., & Pickrell, J. (1995). The formation of false memories. Psychiatric Annals, 25(12), 720-725.

Luus, C., & Wells, G. (1994). The malleability of eyewitness confidence: Co-witness and perseverance effects. Journal of Applied Psychology, 79(5), 714-723.

MacDonald, J. P., & Lavie, N. (2008). Load induced blindness. Journal of Experimental Psychology: Human Perception and Performance, 54(5), 1078-1091. doi:10.1037/0096-1523.34.5.1078

Mack, A., & Rock, I. (1998). Inattentional blindness. Cambridge, MA: MIT Press.

Malpass, R. S., Sporer, S. L., & Koehnken, G. (1996). Conclusion. In S. L.Sporer, R. S.Malpass, & G.Koehnken (Eds.), Psychological issues in eyewitness identification (pp. 295-300). Mahwah, NJ: Lawrence Erlbaum.

Memmert, D. (2006). The effects of eye movements, age, and expertise on inattentional blindness. Consciousness and Cognition: An International Journal, 15(3), 620-627.

Memmert, D., & Furley, P. (2007). 'I spy with my little eye!': Breadth of attention, inattentional blindness, and tactical decision making in team sports. Journal of Sport & Exercise Psychology, 29(3), 365-381.

Most, S.B., & Astur, R.S. (2007). Feature-based attentional set as a cause of traffic accidents. Visual Cognition, 15, 125-132.

Most, S. B., Scholl, B. J., Clifford, E. R., & Simons, D. J. (2005). What you see is what you set: Sustained inattentional blindness and the capture of awareness. Psychological Review, 112, 217-242.

Most, S., Simons, D., Scholl, B., Jimenez, R., Clifford, E., & Chabris, C. (2001). How not to be seen: The contribution of similarity and selective ignoring to sustained inattentional blindness. Psychological Science, 12(1), 9-17.

Neisser, U. (1979). The control of information pickup in selective looking. In A. D. Pick (Ed.), Perception and its development: A tribute to Eleanor J. Gibson (pp. 201-219). Hillsdale, NJ: Erlbaum.

Nelson, D., McEvoy, C., & Schreiber, T. (2004). The University of South Florida free association, rhyme, and word fragment norms. Behavior Research Methods, Instruments & Computers, 56(3), 402-407.

Searcy, J., Bartlett, J. C., & Memon, A. (2000). Influence of post-event narratives, line-up conditions and individual differences on false identification of young and older eyewitnesses. Legal and Criminological Psychology, 5(2), 219-235. doi:10.1348/135532500168100

Simons, D. & Chabris, C. (1999). Gorillas in our midst: Sustained inattentional blindness for dynamic events. Perception, 28, 1059-1074.

Strayer, D. L., Cooper, J. M., & Drews, F. A. (2004). What do drivers fail to see when conversing on a cell phone? In the Proceedings of the 48th Annual Meeting of the Human Factors and Ergonomics Society (pp 2213-2217).

Strayer, D., Drews, F., & Johnston, W. (2003). Cell phone-induced failures of visual attention during simulated driving. Journal of Experimental Psychology: Applied, 9(1), 23-32.

Wade, K. A., Garry, M., Read, J., & Lindsay, S. (2002). A picture is worth a thousand lies: Using false photographs to create false childhood memories. Psychonomic Bulletin & Review, 9(3), 597-603.

White, R., & Davies, A. (2008). Attention set for number: Expectation and perceptual load in inattentional blindness. Journal of Experimental Psychology: Human Perception & Performance, 54, 1092-1107.

Wright, D. B., & Stroud, J. N. (1998). Memory quality and misinformation for peripheral and central objects. Legal and Criminological Psychology, 5(2), 273-286.

Mark G. Rivardo, Kelly A. Brown, Alexis D. Rodgers, Sara V. Maurer, Tyler C. Camaione, Robert M. Minjock, & Gina M. Gowen Saint Vincent College

Author info: Correspondence should be sent to: Dr. Mark G. Rivardo, Department of Psychology, Saint Vincent College, 300 Fraser Purchase Road, Latrobe, PA 15650,
The owner of the purple (white) shopping bag sat at a table and ate
   a sandwich (read a newspaper) right before her friend approached
   her. The thief of the shopping bag was wearing a green (white)
   shirt and was reading a newspaper before she (he) took the bag.

TABLE 1 Frequency of Innattentional Blindness by Conditions

                            Attentional Set

                  Control   Inconsistent   Consistent

Low Expectancy     4/31        25/32         18/30      47/93
High Expectancy    2/31        27/32         20/31      49/94
                   6/62        52/64         38/61
Gale Copyright:
Copyright 2011 Gale, Cengage Learning. All rights reserved.