The purpose of the present study was to examine the multi-tasking behaviors of general
duty police officers, using driving a motor vehicle as the primary task. Fifty-five percent of
the officers were observed performing at least one other task while driving, and 7% of officers were observed using their mobile data terminal (MDT) while driving and handling
another object simultaneously. Our results indicate that the ability to perform the bona fide occupational requirements of a police officer will require the individual to be able to effectively process information, and multi-task. To reduce the performance decrements that occur while multi-tasking, adequate time and attention must be paid to the training of the individual tasks prior to performing them in combination.
Introduction
The word ‘multi-tasking’ implies performing two or more pieces of work simultaneously
such as driving while talking on the radio. Using such a simplistic approach,
it would seem obvious that police officers multi-task during the regular course of
their duties. It would also seem obvious that multi-tasking is not a matter of choice
for police officers, but is imposed upon them by the situational demands of the job.
The demands of some aspects of their work would seem to require that they multitask
frequently and do it with a certain amount of skill. It is difficult to imagine, for
example, how a police officer attempting to control and handcuff a resisting suspect
would not commonly be required to engage in some amount of multi-tasking
(including both physical and verbal control tactics, and inter-officer communication).
It is also difficult to imagine that police officers could simply drive their patrol
cars without ever having to do other tasks at the same time. No doubt, most policeofficers can recall a time while driving at night, when they were looking for an
address, listening and talking on their radio, handling a flashlight or spotlight, and
looking down at their mobile data terminal (MDT—essentially an in car laptop
computer mounted between the driver and passenger seats)—all at the same time or
in rapid succession. Some officers will recall doing all of these tasks while steering
with one knee and typing on their MDT.
Apparently though, not everyone has agreed that police work involves multi-tasking.
In fact, the notion that it does was challenged recently in a lawsuit against a municipal
police department in British Columbia. The lawsuit drew attention to the not so obvious
nature of multi-tasking and the fact that it has not been identified as a bona fide
occupational requirement of general duty police work. Specifically, the lawsuit
involved the claim that while certain activities may look like multi-tasking at first
glance, they are really a group of activities done in succession. The lawsuit went on to
add that it was debatable as to whether or not all parts of a sequence had to be
performed as quickly as police administrators assumed was necessary. The police
department involved was in an awkward situation because, at the time, it had no documentation citing multi-tasking as a bona fide occupational requirement. The lawsuit
was ultimately settled out of court. This suggests that the question of whether or not
police officers are required to multi-task needs to be addressed.
Problem Statement
The purpose of the present study was to investigate the multi-tasking behaviors of
police officers engaged in the execution of their normal duties, and in particular, examine
the use of the MDT in the police cruiser. Observational data were used in the
present study, drawn from a previous data-set examining the physical requirements of
general duty police work (Anderson, Plecas, & Segger, 2001). While that study did not
consider multi-tasking directly, data collected included the number of minutes per
shift officers spent using a MDT while driving and simultaneously performing other
physical tasks. This data allows for an examination of multi-tasking behaviors of police
officers while driving a motor vehicle.
Review of Literature
Increases in technology, and specifically in the area of wireless communications, has
allowed for flexibility in information transmission, particularly as it applies to location.
Telephone conversation, fax transmission, and computer operation are now possible
from the convenience of the motor vehicle. With this increase in spatial flexibility for
communication, many drivers now engage in multi-tasking behaviors, defined here as
performing more than one task simultaneously.
Understanding multi-tasking performance requires an examination of human
information processing, which is based on attentional capacity. While multi-tasking
research is relatively new, attention has been the focus of an extensive research
endeavor since the late 1800s (James, 1890). Multiple definitions of attention haveemerged over this time. For example, Magill (2001) defined attention as ‘the conscious
or non-conscious engagement in perceptual cognitive, and/or motor activities before,
during, and after performing skill’ (p. 117). Cox (1998), in taking a more restrictive
view, defined attention as ‘the product of automatic processes that selectively directs
thinking and behavior without conscious awareness’ (p. 55). In a classical definition,
William James in 1890 defined attention as ‘… the taking possession by the mind, in
clear and vivid form, of one out of what seems several simultaneously possible objects
or trains of thought. Focalization, concentration, of consciousness, are of its essence’
(pp. 403–404). Although not completely in agreement, attention is in common terms
viewed as a capacity or capability that allows information to be processed for the
performance of cognitive and/or motor activities. Critical to the understanding of
multi-tasking behaviors is that attention is viewed as a restricted capacity system that
places limits on human performance.
Theories of Attention
Essentially, two categories of attention theories exist. The first, referred to as central
resource capacity theories, proposes that a single central resource or pool of attention
is available to the individual (e.g., Kahneman, 1973). All tasks would access the same
resource to allow for efficient information processing and effective performance. The
amount of resource available at any one time is flexible and determined by personal,
task, and situational characteristics. Multi-tasking performance is possible as long as
the total resource required for all tasks does not exceed available capacity. A controllable
shift occurs with attention demands to ensure that the ‘best fit’ is found for allocating
the available resource, enabling a discretionary allocation of resources towards
concurrent tasks. Interference between tasks, resulting in performance decrements, is
‘… nonspecific, and depends only on the (combined) demands of both the tasks’
(Kahneman, 1973, p. 11).
The second category of theories is referred to as multiple resource models (e.g.,
Wickens, 1992). In these theories, attention is viewed as a distributed set of resource
pools, each with their own unique capacity and resource–performance relationship. In
Wickens’ (1992) model, resource pools are formed as sub-cells that result from the
interaction of three dimensions—input/output modality (auditory/visual, manual/
vocal), processing codes (spatial/verbal), and stage of information processing (encoding,
central processing, responding). Success of multi-tasking performance results
from the extent to which competing tasks simultaneously tap into the same resource
pool.
According to multiple resources models (Wickens, 1992) the ability to multi-task
depends on the demands placed on attention and which pool the demands originate
from. If the demands come from the same pool, the tasks will be performed less well,
and if the demands come from different pools, task performance is unaffected. For
example, talking and driving can take place simultaneously as they demand different
pools of attention, but performing two motor tasks, such as driving while typing on the
MDT or handling an object (e.g., radio or telephone), is more challenging becausethese tasks demand similar pools of attention. Using this theory, one would predict
more motor vehicle accidents to occur while dialling a cell phone as compared to
merely talking on a cell phone. In fact, Redelmeier and Tibshirani (1997) found that
the risk of being involved in a motor vehicle accident while talking on a cell phone was
four times higher than when not using the phone. This research suggests that performance
decrements result from multi-tasking behavior. These decrements appear to
occur due to limitations in information processing capabilities, not as a result of motor
output interference.
Methods
The analysis of multi-tasking provided here is based on data collected through a
study of the physical requirements of police work. One component of that study
involved full-shift ride-a-longs with 121 randomly selected police officers from all
municipal police departments in British Columbia. During the ride-a-longs
research assistants observed the activities of each officer and recorded those observations
on minute-by-minute tracking sheets. Data were collected on as many as
720 minutes per 12-hour shifts with as many as 49 different activity categories, and
on as many as nine activities per minute. Among the 49 activities observed and
recorded were driving normally and at different code levels, using a MDT, using a
cell phone, talking, handling the radio and other objects, and writing (Anderson
et al., 2001).
For the present study the analysis of data focused on an examination of the percentage
of officers who engaged in other activity simultaneous to using their MDT while
driving. Further, the number of minutes each officer spent during their shift performing
these combined activities was examined, while exploring the differences in MDT
use between male and female officers.
Results
Using ‘MDT use while driving’ as an example, there is no question that police officers
multi-task. Indeed, the analysis of ride-a-long data revealed that the vast majority of
officers (i.e., 77%) were using their MDT while driving. Forty percent used their MDT
while driving on an assigned Code 1 or Code 2 call, and 9% did so while driving backup
to a Code 1 or Code 2 call. As well, officers were observed doing other tasks in the
course of using their MDT while driving. Specifically, 55% were observed doing at least
one other task, and 11% were observed doing at least two other tasks simultaneous to
their using a MDT while driving. Notably, 7% of officers were observed ‘triple pooling’
using their MDT while driving (i.e., drawing from the same neuronal pool as described
previously), while handling another object simultaneously (see Table 1).
Not only did the analysis reveal that the large majority of officers use their MDTs
while driving, it showed that those who do, spend a significant amount of time using
their MDT while driving and performing at least one other task. Specifically, the
analysis showed that these officers spent an average of 17.5 minutes during the shiftusing their MDT while driving. On average, these officers spent 3.5 of that 17.5 minutes
using their MDT while driving to assigned Code 1 or Code 2 calls, and 3.1 minutes driving
back-up to Code 1 or Code 2 calls. Further, officers who used their MDT while driving
spent an average of 4.8 minutes doing at least one other task simultaneously, while
spending 1.5 minutes doing at least two other tasks, and 1.8 minutes performing three
tasks that require attention resources from the same pool—triple pooling (see Table 2).
Differences between male and female officers’ use of their MDTs were observed. As
Table 1 indicates, while there were no differences between male and female officers in
terms of the percentage of each who used their MDT during the shift and while driving,
a significantly greater percentage of female officers used their MDT while driving backup
to Code 1 or Code 2 calls. Further, a significantly greater percentage of female officers
were observed doing at least two other tasks simultaneous to their using their MDT
while driving and nearly 24% of them (vs. only 3% of male officers) were observed
‘triple pooling.’ Further, female officers who used their MDTs during their shift, did so
for a significantly greater number of minutes. Specifically, female officers used their
MDT on average for 79.7 minutes during the shift while male officers did so for 49.3
minutes. As Table 2 indicates, female officers also spent a significantly greater time
using their MDT while driving.
One of the most interesting findings of the analysis was that use of MDTs appears to
be something that is more often than not associated with multi-tasking, whether driving
or parked. Specifically, as Table 3 shows, officers were multi-tasking 84% of the
time they were using their MDTs. That is, simultaneous to using their MDT, they were
either writing, handling an object, driving, using a cell phone or their radio, or talking
to a suspect or another officer (or any combination of these activities). As the table
shows, 32% of the time they were driving while using their MDT and 5% of the timethey were doing this and something else. Less than one-half of 1% of the time they
were doing two other activities and equally as infrequently, they were triple pooling
using their MDTs while driving. Still, the point is that on average they do multi-task at
very sophisticated levels at least some percentage of the time.
Discussion
Research demonstrates that humans are capable of multi-tasking behaviors, but the
degree of performance success is related to the available resource for attention and theattention costs of the activities. Resources and costs are determined by a complex array
of factors, notably arousal levels, age, and stage of learning. Police officers, due to the
nature of their occupation, would appear to have the need to multi-task frequently
during the course of their work shifts. Success at this activity would result from the
officer’s attentional resource capacity coupled with the attentional costs of the activities
that are being performed. Theories on attention and attentional capacity have examined
the restricted nature of attention. Some theories suggested a bottleneck in the
information processing system (Abernethy, 2001; Logan & Gordon, 2001) or a limited
availability of resources (Logan & Gordon, 2001). Wickens (1992) proposed a multiple-
resource theory of attention with three distinct resource pools that include input/
output modality (auditory/visual, manual/vocal), processing codes (spatial/verbal),
and stage of information processing (encoding, central processing, responding). The
allocation of resources to each system would depend on the required task (Abernethy,
2001; Magill, 2001). Being able to multi-task would depend on the nature of the
demands of attention and the pool the demands originate from. Performance decrements
during multi-taking would occur when the attention demands were from the
same pool, as can be demonstrated by the difficulty in simultaneously rubbing your
head and patting your stomach. Further limitations to multi-tasking are imposed by
high levels of arousal, affecting the ability to attend to the vital information being
presented in the environment.
Applying Wickens’ (1992) theory to the present data would suggest that 55% of the
officers observed could be expected to have performance decrements as were observed
driving and using the MDT while performing at least one other task. Further, 11% of
the officers were observed doing at least two other tasks simultaneous to their using a
MDT while driving. From this data, 7% of officers were found to perform three tasks
drawing from the same pool of resources, using their MDT while driving while handling
another object simultaneously. It is this group of officers which would be expected to
have the largest performance decrement in any (or all three) of the tasks being performed.
Regardless of the theoretical framework, several factors have been proposed that may
modify the attention–performance relationship and impact on successful multitasking
behaviors. A number of studies using both learning (e.g., Damos, Bittner,
Kennedy, & Harbeson, 1981; Spelke, Hirst, & Neisser, 1976) and expert/novice
(Leavitt, 1979; Parker, 1981; Smith & Chamberlin, 1992) paradigms have indicated that
multi-task performance can improve with practice. It is not exactly clear why this
occurs. Several possibilities have been suggested in the literature, and include:
1. a reduction in resource needed to perform one, or more, of the multi-tasking activities
as a result of a shift in control processes from conscious to automatic processing
(e.g., Schneider & Shiffron, 1977; Shiffron & Schneider, 1977);
2. maximizing resource pool availability through optimal arousal control (Kahneman,
1973);
3. more effective information processing behavior, which may include time-sharing
or attentional switching strategies, or spreading the processing requirements across
different resources (e.g., Allport, 1980).
Although learning results in improved multi-tasking behavior, it has been pointed
out that learning, as a process, is attention demanding and that engaging in multi-tasking
behavior during learning can be detrimental to subsequent performance levels. In
particular, a division of attention during the encoding process in memory formation
can result in negative performance artifacts (Musen & Viola, 2000; Naveh-Benjamin,
Craik, Perreta, & Tonev, 2000; Salthouse, Fristoe, Lineweaver, & Coon, 1995; Schmitter-
Edgecombe, 1999). Therefore, if subjects are engaged in the execution of a concurrent,
attention-demanding task when items are available for inspection and encoding,
their subsequent performance on a task will be much lower than if full attention had
been given to that task while learning (Isingrini, Vazou, & Leroy, 1995). This implies
that tasks that must be performed in a combined manner should be acquired in isolation
initially, perhaps to the point of automaticity, before performing and practicing
tasks concurrently. If subjects are engaged in a concurrent attention-demanding task
when items are available for inspection and encoding, their subsequent performance
on a task will be much lower than if full attention had been given while learning (Isingrini
et al., 1995). This would suggest that both driving and MDT use should be
mastered separately before being performed in combination. All mechanical operations
(driving, MDT, and radio use) should be mastered in isolation with automated
responses that require little attention resource allocation. Once automated, the tasks
should be gradually introduced in combination, and practiced in the performance
environment.
The boundaries for attentional resources required for optimal performance depends
on a number of factors. For example, if the task is a newly acquired or difficult task, the
boundaries for attentional capacity are smaller, and the task performance will be
poorer if boundaries are stressed. A well-learned or simple task, on the other hand,
would allow for more flexible boundaries (Abernethy, 2001). Practice allows the organization,
shaping, and reduction of the attentional demands of one or both tasks (by
processing the task automatically), the development of new time-sharing and attentional
switching strategies (that reduce inter-task interference, and produces a more
economical way of functioning), and increase the availability of capacity or resources
through optimization of arousal. For this reason, adequate time and effort must go into
recruit training, even for the simplest of tasks (e.g., handling the radio). The introduction
of a new technology or procedure without adequate training experience may result
in placing the police officer at risk during the execution of their duties.
Summary
A number of theories have been proposed to explain the application of attention to
skilled behavior; in general, attention has been identified as a resource that can be
applied to information processing. Performing a task effectively would have a particular
attention cost associated with that task. If the resource available to the individual is sufficient
to meet the costs of performance, then task performance should result in an optimal
level parallel to the individual’s capabilities. However, if the attention cost of
performance exceeds available resource, performance will suffer. This paper demonstrates, using driving as the base activity, that 77% of police officers
observed performed multi-tasking during their shift. Further, 55% were found to
drive, use the MDT, and perform at least one other task simultaneously for on average
4.8 minutes a shift. Of significance, 7% of the officers performed multi-tasking behaviors
that required the use of the same attentional resources during which a performance
decrement in any one task (or all three) is expected.
Conclusion
The ability to perform the bona fide occupational requirements of a police officer will
require the individual to be able to effectively process information, and multi-task. The
present research demonstrates the need to pay attention to the multi-tasking requirements
of policing, specifically those occurring during driving. The demands on the
motor control of multiple tasks put strain on one pool of attention resources and may
lead to performance decrements. To reduce the decrements that occur, adequate time
and attention must be paid to the training of the individual tasks prior to performing
them in combination. This will have implications for the training of new recruits and
incumbents who are introduced to new technologies.