Page 23 - JSOM Fall 2023

FIGURE 1 Examples of the drift diffusion modeling (DDM) evidence to support a link between shoot/don’t-shoot errors
process and how it can provide refined detail about shoot/don’t- and inhibitory control failures, but the link appears to depend
shoot decisions. more upon volitional stopping capability than automatic inhi-
bition. 34–38 In essence, failures depend upon both the speed of
decision-making processes and the individual ability to stop—
the gas pedal and brakes, respectively, of this cognitive car. Un-
fortunately, static go/no-go tasks are suboptimal in exploring
volitional stopping power. 39-41 The better measure involves a
task with some element of dynamic change, such as a go sig-
nal that suddenly switches to a no-go signal. There are many
38
ways the task can be modified to incorporate this component,
but in short, some change or dynamic element should allow
the stimuli to occasionally present as one category and switch
to a different intended designation.

This task supports optimization in multiple ways. Foremost,
information processing speed can be quantified, and this de-
cision parameter represents an ideal operationally relevant
component to enhance. Faster information processing speed is
crucial during the high-pressure scenarios encountered by Spe-
cial Operations Forces (SOF). Accuracy and inhibitory control
lower threshold make the Operator more prone to decision er- become further variables to optimize as a means of identifying
rors since evidence accumulation in this context is a stochastic and enhancing individual impulsive behavior. Furthermore, in-
process. Errors occur when information accumulation reaches formation processing can be distinguished quantitatively from
the wrong threshold, which in practice is a quantitative way response speed in a way that permits further scrutiny and tar-
to model a threat assessment error that could result in firing geted interventions. A cognitive enhancement practitioner could
upon an ally or civilian. In turn, despite the overall response determine whether an individual is making errors because he
speed being similar between the two Operators, one individual or she is not processing information fast enough, making snap
can clearly be identified as the superior decision-maker—all of judgments based on inadequate decision thresholds, inhibitory
which can be modeled from a decision that occurs in less than control failures, or possibly even errant motor activity. Each is-
a second. sue would benefit from a different cognitive enhancement tech-
nique, and DDM calculations provide the precision to support
In this example, the upper boundary represents a threat de- optimization through more targeted enhancement. Thus, the
cision and the lower boundary represents a non-threat deci- implication is that a more precise test can enable more precise
sion. The four DDM parameters as modeled are drift rate (δ), follow-up interventions should a problem be discovered.
which signifies the amount of information uptake during the
decision-making process; decision bias (β), which signifies Speed, the final Special Operations cognitive requirement, is
whether an individual is biased to make a threat or a non- also satisfied with this threat/non-threat task. Although the
threat decision; decision threshold (α), which signifies how DDM calculations benefit from more trials, the decision pa-
much information an individual needs to reach a decision; and, rameters can be modeled from several dozen trials. 42,43 This
non-decisional processes (τ), which includes factors such as trial requirement means that the task could be completed in
motor response speed. as little as 5 to 10 minutes. A minimal time requirement to
conduct the assessment supports a smaller logistical footprint
Previous research has used the DDM approach to model and allows for multiple assessments or more Operators to be
shoot/don’t-shoot decisions, which provides prior empirical assessed in a shorter time course. More trials would allow for
success and examples on which to build. 26–29 A major theo- more stimulus variety and better exploration of individual ad-
retical and practical concern becomes whether to conceptual- herence to assigned rules of engagement, but the task can be
ize the response a single button-press, as in go/no-go tasks, or completed in as brief a timeframe as 5 minutes per Operator.
the more commonly used two-alternative forced choice task,
which requires participants to press one of two buttons on These combined qualities suggest that the task would be far
any given trial. It is somewhat counterintuitive to use a sin- better for Special Operations cognitive assessments than those
gle-response paradigm for DDM calculations because identi- cognitive tasks currently employed by the ANAM and other
fying characteristics associated with the alternative decision cognitive batteries. The base task is a dynamic threat/non-threat
boundary requires some response to be made. Even so, recent assessment task supported by DDM calculations to identify in-
research has demonstrated that DDM can successfully be cal- dividual capabilities in processing speed, decision thresholds,
culated using a single response threshold and implicit lower and inhibitory control capability. The remaining discussion will
response boundary where no response is actually made. 30–32 shift to the research and development requirements needed to
This single-response version can better resemble the classic go/ support this type of cognitive modeling in the force.
no-go task, and function as an inhibitory control assessment,
as well by identifying the number of commission errors made Next Steps to Support a Special Operations
during the threat assessment. 33
Cognitive Assessment
One further modification can be made to better support the Despite the potential in using a dynamic threat assessment
task as a measure of inhibitory control. Specifically, there is to measure cognitive function, there remain several logistical

SOF Cognitive Assessments | 21