Page 3 of 4


The Attention Study: A Study on the Presence

of Selective Attention in Firearms Officers


Again, note that the pattern of the constables’ scores regardless of the number of items listed is overwhelmingly large in the External category versus the Internal category and exceptionally higher by far in the External Narrow as compared to the External Broad.


Comparison Between Conferring and Nonconferring in the Interview Category


As with the report groups, generally the researchers would not have expected to have a difference in the Internal category between these two groups in the Interview categories. This is because this Internal quadrant relates to the constables’ internal thoughts, reflections, and awareness of their own behavior.


Considering that most of the constables in this condition also did not mention their thoughts or make comments about their own behavior during the conferring and that other constables did not elicit their thoughts, there should not be a difference in the Internal category between Conditions III and IV. If there was any expected increase, it should have occurred in the External category because of the conferring. However, during the conferring process, the researchers observed here as well that the constables did not appear to learn anything “new” during the conferring. Therefore, it is not surprising that there was not a difference between the information reported by the constables in any of the quadrants in the Interview condition from those constables who conferred and those who did not confer.


The most significant difference between the conferring and nonconferring interviewed groups occurred in the number of errors that each group made. This is explained more completely in the “Error Rates” section. It is important to note here that in the External Narrow quadrant (which is the quadrant in which the constables would be noting the behavior of the subject, the threat presented, etc.), the group that did not confer had 25% more errors than the group that conferred.


Narrow Versus Broad Focus of Attention


When the averages were combined across both the narrow and broad focus of attentions, the average across all of the experimental groups (conferred, nonconferred, interviews, reports) was four times larger in the External than in the Internal categories. The average for the External focus was 55.31 items noted versus 13.47 items noted for an Internal focus.




When the scores are combined across all of the conditions and compared to a Narrow versus Broad focus of attention, the Narrow focus of attention had an average of 44.77 items recorded for the average constable. The Broad focus of attention had 24.01 items recorded. This was true even though the graders were very liberal in putting External items in the Broad category. Therefore, as supported by the data, it is overwhelmingly clear that the constables had a Narrow focus of attention.


This means that when the constables were engaged in the performance of assessing and reacting to the scenario and in engaging the hostage taker with their firearm, they had an External and a Narrow focus of attention. For the most part, because of the close proximity of the encounter, the constables were not using their sights but were using a kinesthetic alignment (point shooting). Although this method of shooting did not require sight alignment, it still occupied some portion of the constables’ attentional processes.


The important element about an External focus of attention, particularly an External and a Narrow focus of attention as most frequently occurred in our study, is that the process allows the constables to focus on what is important to them at the time. Unfortunately, they then can miss other items that later may turn out to be important. For example, if the constable’s attentional processes at the moment of the shooting were on the alignment of their gun on the hostage taker’s center mass, they then did not note anything about or paid minimal attention to anything else about the subject at that moment. This includes the specific movement of the subject, the subject’s clothing, the subject’s actions toward the hostage, etc. Under this condition, they would note that the subject had moved, for example, but be unable to explain the precise frame-by-frame movement of the hostage taker as they were simultaneously engaged in shooting.


This matches the research results found in the attentional process of successful athletes. Wulf (2007) notes that an External focus of attention is a primary prerequisite for successful performance in an athletic competition. Vickers (2007) adds that an External and Narrow focus is vital for successful performance of a psychomotor skill in athletics, from golf to basketball to football. It is reasonable that trained firearms officers will function in a similar psychomotor fashion to successful athletes when the officers are performing a psychomotor skill.


Therefore, the data is overwhelmingly clear that the constables had an External Narrow focus of attention during the most stressful component of the scenario, and this appeared to facilitate their performance and effectiveness but simultaneously impaired their ability to provide full and complete reports about the incident.


The use of Nideffer’s Quadrants of Attention (Nideffer & Sharpe, 1978) was overwhelmingly supported as being an appropriate tool for categorizing and understanding a constable’s focus of attention during a high-stress force encounter. To use law enforcement terminology, the research overwhelmingly supported the constables’ experience of tunnel vision and tunnel hearing or perceptual distortions during this type of encounter.


The results of the research also clearly illustrate that the constables’ focus of attention in this type of encounter was primarily external in both the External Narrow and External Broad modes. The constables who were interviewed had four times more observations in the External than in the Internal quadrant. The constables who wrote reports had over six times more information in the External quadrants than the Internal quadrants.


Even though the constables were more focused on the External quadrants, they were more narrowly focused than broadly focused. Even when they were broadly focused, they still were narrowly focused on a particular person or event in the environment. This was in spite of the graders being very generous in grading the External Broad category. For example, if the constable noted the shotgun assailant was wearing a shirt, the constable would get a score in the External Broad category. If they noted the shirt had any particular features, such as short sleeves, then the constable would get a score in the External Narrow category.


As noted, an External and in particular an External Narrow focus of attention significantly facilitates a focus of attention on a threat and, in turn, facilitates great performance by the constable; however, it also renders the constable attentionally blind to anything that they are not focused upon. To paraphrase one constable, “I knew what I saw, but I don’t know what I didn’t see.”


Also, the constables’ accuracy of their reported memory waned the further they were away from the details on which they were specifically focused. For example, every constable who could see the shotgun usually reported a quite detailed and accurate description of the shotgun. They were accurate about the shooter’s action while they were focused on it. They were less accurate about the clothing the shooter was wearing. They were often quite unobservant or inaccurate about the hostage, and they almost never noted the behavior or action of anyone other than the shooter. Further, they almost never noted the presence or action of the shooter’s brother and even other constables—unless, of course, that constable or brother was somehow directly involved with them such as being directly between them and the assailant with the shotgun. Subsequently, we can state that the constables’ focus of attention was very narrowly driven and externally specific. However, depending upon the constable, the constable’s location, what the constable was attempting to do, and so on, the constable’s specific focus varied and evolved during the incident and was often unique to that constable and the behavior on which that constable was focused and/or on what he or she was attempting to accomplish.


Thus, we can definitely state that the constables in this study had a very selective attention that was driven by the evolving incidents in the scenario and the constables’ own attempts to respond to those events. They were attentionally blind to anything on which they were not focused.


This means that the constables made errors in reporting items, particularly items or behavior that occurred on the periphery of their attentional focus. We also recorded their errors.


Error Rates


As shown, the constables were both accurate and inaccurate in regards to their recollections about their incident from the moment of the assault by the hostage taker up to the constables’ control of the incident. These error rates during that portion of their incident will be discussed in this section. However, before the data in this portion is discussed, it is important to note that the constables were also both accurate and inaccurate in their reports on the scenario before the assault by the hostage taker. For instance, the constables did not note or report on the relationship between the individual engaged in the confrontation at the front desk and the wounded bank robber. In the scenario created for this research, this relationship was not important; however, it was also not noted by the constables during the initial conflict at the desk.


The results of the analysis of the error rates—that is, the number of items the constables incorrectly reported on—is truly astounding and definitely needs to be investigated further.


The following error rates were observed across the only quadrants that error rates could objectively be assessed in: External Narrow and External Broad. Under these conditions, the constables noted some particular behavior or items in their reports or interviews, and the graders could then assess whether or not that behavior actually occurred and was documented by video from one or more of the three video cameras recording the constable’s scenario.


The following tables present the error rate data grouped by the two viable foci of attention and the four treatment groups. Table 8 simply depicts the average number of errors for each constable in each category. Table 9 combines both the averages for the error rates for each constable in each category and the average number of items the constables reported in each of the categories. By combining them in this way, it is possible to assess the volume of items reported and the error rate in relation to that volume.



The reader will note that the lowest error rates were recorded in the Report categories by the constables who wrote reports. The average number of errors for each constable in this category was somewhat less than half of one error per incident per constable. The errors were 0.45 per constable in the External Broad for the nonconferred/report condition, 0.50 per constable in the External Broad conferred/report category, 0.60 per constable in the External Narrow for the nonconferred/report condition, and an astounding 0.14 per constable for the External Narrow category for those constables who conferred and wrote reports.


This may be a product of the small amount of information provided in the constables’ written reports. The smaller amount of information provided by the constables in this category meant the constables had less chance of being mistaken.


The notable exception in the Report category was the exceptionally small number of errors in the conferred/report category for the External Narrow quadrant of attention. Fourteen constables were included in this group, and each constable reported 22.43 items or behaviors in this quadrant. The average error rate for every 22.43 items was 0.14. It is an amazing statistic. Framed another way, 14 constables provided a total of 314 correct details (14 constables ⋅ 22.43 items) in this category and only recorded two factual errors in all of that data.


The smaller number of items reported in this category cannot be the only reason for this incredible statistic. The constables in the nonconferred/report group had about as many items noted but had four times the error rate in this category (0.60). This is still a small number, but it is about four times greater than the error rate of the constables who conferred and wrote reports.


Also, those constables who were interviewed, although they provided, on average, more than twice the amount of information in the External quadrants, they had an error rate for that information that was very high. Although the difference under the interview conditions between those constables who conferred and those who did not confer is not as dramatic as the difference in the report writing conditions, the constables who conferred and then were interviewed made 25% fewer errors in the External Narrow quadrant than the constables who did not confer and then were interviewed. This means the constables who conferred and then were interviewed were considerably more accurate in reporting those behaviors that they were narrowly focused or tunneled in on than the constables who did not confer and were interviewed.

The noticeable comparison is on the error rate in the “External Narrow” quadrant between those constables who conferred and wrote a report and those who did not confer and were interviewed. Those who conferred and wrote reports provided about a third of the information about that which they were “Externally” and “Narrowly” focused on as compared to the information provided by those who did not confer and were interviewed. However, those who conferred and wrote reports had fewer errors by volumes in this quadrant than did the constables who did not confer and were interviewed. The constables who conferred and wrote a report had an error rate of 0.14 per constable. This means when all the errors are combined across all of the constables in this category, only two errors were made by all of the 14 constables in all of the information they provided about that which they were externally and narrowly focused on.


In comparison, the constables who did not confer and were interviewed had an average of 6.6 errors per constable in this category in the External Narrow quadrant. The constables in this category reported a total of 339 items (56.5 items per constable ⋅ 6 constables) and had 40 errors in total. There were only six constables in this category, and although statistical analysis not provided in this report indicates the data is still reliable, due to the small sample size, the author of this paper is still suspicious. The error rate for those six constables who did not confer and were interviewed is 47 times that of those who conferred and wrote reports.


Another interesting result was found when comparing the error rate of those who conferred and wrote reports with those who conferred and were interviewed. Fifteen constables were in the confer and interview group. This is a significant size difference from the six constables in the nonconfer and interview group and should result in a good comparison. Each constable in the confer and interview group reported an average of 51 items in the External Narrow quadrant. This is a total of 765 items (51 items ⋅ 15 constables), and they had a total of 66 errors compared to the two errors for those 14 constables who conferred and wrote reports.

Regardless of whether the constables conferred or not, the interviews led to a significant number of inaccuracies in the information that the constables reported. It is the observation of the graders and the author of this paper that the primary source of the errors dealt with information on the edge of the constables’ focus of attention. The constables worked hard to provide accurate information, but the interviews apparently led them to expand on items that they were less knowledgeable about.




A very credible conclusion from the results of this research might be that interviews produced significantly more data particularly in regard to the items and behavior that the constables were focused on during the incident, but those who were interviewed also had a significantly higher error rate on that increased quantity of information. The written reports produced less information but produced fewer errors and, therefore, were much more accurate in their totality. Conferring, as was done in our study, produced fewer errors around that information on which the constables conferred, which was primarily items and behavior in an External Narrow focus of activity and attention. Therefore, a recommendation from the results of this study would be that when constables are going to have to report on an incident, the most accurate reporting of the details is going to be provided by those constables who have had a chance to confer and write a report. A further conclusion might be that the fewer details the constables provide about the incident, the fewer errors they will make especially if they report only on the behavior that they were most specifically focused on.


This study and others, such as Just et al.’s (2001), inform us of the value of a single, undivided attentional focus. They also inform us of the liability of this same process, which results in, among other things, inattentional blindness to anything the constable is not focused on.


LeDoux (2002) and Damasio (1994) inform us as well that when human beings are performing under a high level of threat, they lose higher cognitive functioning which includes their executive functions such as the ability to critically think and perform logical analysis.The scenario created in this study caused a high level of stress in the constables, as measured by their self-reports and the scores from heart monitors. This resulted in a significant narrowing of their attention to the hostage taker/assailant and his shotgun and their response to that threat. This narrow focus of attention toward the rapidly evolving and very threatening situation occupied all of the constables’ attentional processes. This narrow focus of attention facilitated the constables’ perception of and response to the threat, but it also created a significant cognitive impairment.


It is extremely difficult, if not impossible, for the constables or any human to be so intently and emotionally engaged in a scenario such as this one for such a brief period of time and also to simultaneously be critically assessing all of the related elements in the scenario as well as their own behavior. For instance, most of the constables noted that they heard the shotgun being fired, and many noted that it had been fired twice before it had been pointed at them. They also noted as it was being pointed at them that it was a double-barreled shotgun.


Only one team of three constables out of a total of 46 constables was able to note that the hostage taker had fired both barrels from the double-barreled shotgun and, subsequently, it was no longer loaded with viable rounds. That group noted this after they had attempted to verbally control the hostage taker who was pointing the shotgun directly at them. When the hostage taker, now with enough time, attempted to reload his gun, these constables saw that it was empty and rushed to control him.


This was not a behavior that the researchers were looking for, but it is consistent with the focus and the finding of this study. That is, that human beings attempting to perform in a high-stress incident of short duration lack the attentional processes and cognitive functioning to engage in life-saving behavior while simultaneously critically analyzing that behavior. Given the opportunity and time of the course, this process could occur as it did in a small number of participants in our study.