Académique Documents
Professionnel Documents
Culture Documents
Combat is inherently a demanding situation which may increase stress, heighten arousal,
and increase anxiety. The Processing Efficiency Theory (PET) was specifically
developed to account for how anxiety influences performance. Therefore, the PET
provides a potential theory to explain the positive and negative changes in performance in
a combat environment. This study is the first attempt to examine PET in the military
domain. Using the Small Arms Simulator Testbed (SAST), we investigated the
relationship between processing efficiency and performance, considering mental
workload, stress, and anxiety effects. Shooting performance effectiveness measures were
target acquisition and friend/foe discrimination. Changes in processing efficiency were
manipulated by varying degrees of working memory demand and sustained information
transfer. The results indicated that shooting performance and processing efficiency, as
well as mental workload demands, decreased as the global demand of both tasks
increased. Further analyses for anxiety and stress and future directions are discussed.
of Central Florida and were paid $22.50 for their pressed in response to the presentation of a valid cue
participation (based on an hourly rate of $7.50). word and true association word, the second trigger was
pressed in response to the presentation of a valid cue
Materials word and false association word, and the third trigger
was pressed in response to an invalid cue word.
Questionnaires. Participants completed several
subjective questionnaires to measure individual Apparatus
differences including workload, stress, and anxiety. All
subjective measures were administered via Inquisit, Small Arms Simulator Testbed II (SAST II).
psychological data collection software, on a Dell (Naval Air Warfare Center Training Systems
Inspiron 5150. Division, 2000) is a single user research testbed for
the evaluation and study of both current and new
Experimental Tasks weapon concepts. The SAST controlled the
Primary Task. Target acquisition task and a
presentation of targets (target type, location, time of
Friend/foe discrimination task: two different types of onset, and exposure time), and recorded weapon
stimuli/targets appeared on the screen (Figure 1): the aim-point data and shot results from an M16 rifle
non-target was an M-16 (friend), the target was an AK- equipped with an infrared emitting diode and
47 (foe); both stimuli/targets were superimposed over collimator lens. Weapon recoil was simulated with
the silhouette stimulus; the discrimination was based an electromechanical recoil system and weapon
upon the weapon. The spatial location of each stimulus sounds were simulated via digital audio surround-
(targets/non-targets) was randomly generated to prevent sound (as shown in Figure 2).
the predictability of the appearance of a target.
AK47
M16
stimuli (target, or non-target) and in half of the trials, the Working Memory Demand
target present than when a target was present and when DISCUSSION
the working memory demand level was two (see Figure
4). This study is the first to investigate processing
efficiency in a military context and in a high fidelity
95 simulator. The use of a high fidelity simulator is
93 important in this case because the target aiming and
91 target acquisition task are performed using a real weapon
89
which adds to the physical and the hand-eye
Percent Corre
87
Target coordination demands of the task. As can be seen from
85
No target the results there are changes in performance for both the
83
primary and the secondary task as the global demands
81
change, clearly indicating that it is not the case of one
79
task dominating the other. However, contrary to our
77
hypothesis, when the task was most demanding (both
75
2 4
primary and secondary tasks were present in the highest
working memory demand condition), participants were
Working Memory Demand
more accurate on the shooting task when the secondary
was present than when there was no secondary task. One
Figure 4. Relationship between the presence of a target
possible explanation for these results is that there were
and working memory demand.
combined practice and fatigue effects. Hence, the
variability among participants increased, and possibly
Perceived Workload. A 3(working memory demand)
the analysis of the anxiety data could contribute to the
X 4(block) within-participants repeated measures
understanding of these changes. Additionally, because
ANOVA was conducted on the subjective ratings of
participants were significantly less accurate on the
mental workload (RSME). There was a significant main
recognition task in the same high demand condition, it is
effect of working memory demand and block, Wilks
possible that participants stopped attending to the
= .35, F (2, 31) = 29.32, p = .001, partial 2 = .65, and
secondary task and focused all of their attention on the
Wilks = .55, F(3, 31) = 8.08, p = .001, partial 2 =
primary, more salient shooting task. This would be
.45, respectively. Specifically, participants reported that
consistent with the Hancock and Warm (1989) model
a higher level of mental effort was required as a function
indicating that participants were no longer in their
of working memory demand; as the working memory
comfort zone and were unable to maintain an optimal
demand increased, the ratings of mental effort increased
level of performance on both tasks. Furthermore, such
(see Figure 5). Additionally, and consistent with our
pattern could also be explained using Hockeys (1997)
hypotheses, participants reported that a lower level of
compensatory control model. Specifically, it may be that
mental effort was required as a function of time in each
as attentional demand increased participants regulated
working memory demand condition.
their effort downward on the secondary task to free
75
resources required to successfully engage the primary
task.
65 Furthermore, the mental workload measures
Block 1 correlate with this global demand, indicating a
Correct
55
Block 2 performance-workload association (Hancock, 1996; Yeh
45 Block 3 & Wickens, 1988). Further investigations of this
Block 4 relationship are necessary. Current efforts are underway
35 to examine the hypothesized effects of trait anxiety on
performance and perceived workload. In addition, the
25
0 2 4
other form of workload assessment, physiological
measurement, will be examined to test whether there is
Working Memory Demand
physiological evidence that participants reduced their
effort in the face of high working memory demand.
Figure 5. Relationship between subjective mental effort These analyses will provide a more coherent
and working memory demand. understanding of the results. Future studies will be able
to refine these findings and better explain the tradeoff in
performance as task demand increases.
PROCEEDINGS of the HUMAN FACTORS AND ERGONOMICS SOCIETY 49th ANNUAL MEETING2005 2123
ACKNOWLEDGEMENTS
REFERENCES