In response to the following missive I sent to this forum
>All aviation training is about instilling behaviours
>in aircrew that will result in timely and
>appropriate decision/actions...
>
>...It argues for a skill based approach to CRM,
>which after all is what training is generally all about.
>Modification of crew's attitudes is difficult at best and
>the vote is still out on the success of this
>approach in modifying behaviour...
====== end snip ========
John Wise wrote
To my biased way of thinking - behavior is the only thing worth
measuring in a CRM environment. If we were not trying to change
behavior - why bother with CRM. Is it is considered a general ed
class? I am interested in the work that Keith discussed taking
place at DCIEM. Perhaps he could give us a little more information.
_________________________oOOo______________________
This exchange is now rather old now and I apologise to John, and perhaps
others,
for not replying in a timely manner. Now I have a little more time on my
hands I
thought I would try to resurrect this topic and respond to JohnÕs request for
information.
Over the last 10 years we have been doing work at DCIEM aimed at
understanding the concept of operator workload to the point where we
can develop empirical models that could be used in front-end human engineering
design tools. As new paradigms came along (Situation Awareness, Cognitive
Compatibility etc.) these were absorbed into our thinking.
A review of the literature on workload shows relatively little agreement
except on
the following two statements
¥ workload is multi-dimensional, and
¥ workload appears to have a time-based component, an intensity-based
component (related to complexity, difficulty etc.) and a psychological stress-
based component.
We ran a number of experiments in which we tried to manipulate time stress
and intensity (in the sense of the amount of information to be processed)
separately. The experimental environment involved an ATC simulation.
Screen update time (this sets the allowable time for control actions) and
number of aircraft were manipulated independently. What emerged out of all
this was a model of workload (the IP Model) that states that performance,
errors, and estimates of workload all depend on the time pressure (TP) as
given by the ratio
TP = (time to respond) divided by (time available before action has to be
taken)
Time to respond is directly related to the amount of information to be
processed
(this model is based on a simple application of information theory). Hence
all
factors that influence workload can be reduced to their effect on the amount
of
information to be processed or their effect on the time available before the
action
has to be taken. A paper describing the development of this model appeared
recently
in Human Factors 39(1), pp30-47.
To this model we have added aspects of Perceptual Control Theory (PCT) which
makes the claim that all human behaviour occurs as a result of the closed loop
control of oneÕs perceptions (in other words we act on the world to change
itÕs
state so that our perceptions match desired goal states - our goals are the
set
points of the loop). This is a classical control theory model, so all the
rigour
of this discipline can be brought to bear. Together the IP Model and PCT
brings into one conceptual framework much of what we know about performance,
error production, workload, and SA Ñ plus the interactions between them.
A paper dealing with these issues is on the CRM developers web site at
ftp://www.caar.db.erau.edu/pub/crm/hendy.pdf
As well as satisfying our original aim to provide an empirical model of
operator
workload that could be used in predictive methods for design, the IP/PCT
Model has provided considerable insight into things like team performance,
CRM and human interaction with automation.
Behaviours are the manifestations of decisions that have been
made (the PCT Model) and these decisions are the results of processing
information (the IP Model). The IP/PCT Model is therefore a decision making
model. It would suggest that if we wish to understand behaviours (including
training
them, assessing them and bounding them) we need to understand what the
decision
makerÕs goals are (the set point) and their state of knowledge of the world
(including
themselves). This internal knowledge state includes what many would call SA,
plus all the other knowledge we have about the world and our place in it (the
mental model). It seems obvious to me that not only do we need to focus
on behaviours but we need to look at the decisions behind these actions. At
the end of the day it is the actions (including the decision to do nothing)
one
takes that determines the safety and efficiency of a flight.
Recently, to address some problems seen in our C-130 operations, we have
used the IP/PCT Model to build a draft syllabus for a new training
package ÒHuman Factors in Decision Making.Ó We would envisage that this
package would replace what has been traditionally dealt with as Aircrew
Co-ordination Training (aka CRM). Many of the CRM topics will be covered
under the proposed training, but all topics will be tied to how they effect
decision making rather than presented in isolation. Using the IP/PCT Model
the
course topics become evident as does the desired knowledge, observable
behaviours and the method of assessment. A paper describing this effort
was presented at the Aviation Psychology Conference in April. I can have
it put on the web site if there is sufficient interest.
As a collateral study we are also addressing the issue of captaincy. Through
a formal knowledge elicitation process we are trying to define the dimensions
of the observable behaviours associated with good, versus bad, captaincy.
We are looking at this both from the perspective of aircraft commanders and
the rest of the crew (by position). I tried to get some interest going once
before on this issue but it fell on barren ground :-). I have attached some
material
covering this work below - it has already been distributed to a couple of
people
on this forum to try to get some involvement.
_________________________oOOo__________________________
We have identified two problem areas with our CC-130 community
¥ lack of systems knowledge
¥ weakness in Captaincy/Decision making skills
Obviously this only applies to a small group in the organisation and is
possibly largely due to lowering experience levels (down from a mean
5000-6000h for an AC 15 years ago to 2200-2500h). Addressing systems
knowledge is relatively straight forward but the issue of Captaincy is more
difficult.
As you know from OSU we are attempting to re-orientate the ACT/CRM course
to a *HF in Decision Making* course based on the framework I presented in
Columbus. I have a draft syllabus that our Air Transport Group are
considering.
In each module of the course we are trying to identify the take away knowledge
and the demonstrable skills that should result. But what are the behaviours
associated with good Captaincy?
To address this last issue we are conducting a formal knowledge elicitation
exercise (the Rep Grid) with CC-130 crews. The goal is to tease out the
observable behaviours crews associate with good versus bad Captaincy. It
takes
1 to 1.5 hours per person (we have done over 70 now). The program runs on a
Macintosh computer and we have written full operating instructions on how to
administer the procedure.
Though personal contact and through more formal US-AS-UK-CA agreements I have
been trying to get other airforces to participate - so far without much
success.
I think I can convince the ASs (and maybe the Brits) to get involved at a
forthcoming meeting of TTCP (a scientific collaboration agreement between the
Anglophone countries).
Do you have any contacts in the CC-130 world (or potentially any fleet) that
might see an advantage in doing such an exercise. In both the military and
the
Civilian world there seems to be a poor understanding amongst some of what
Captaincy is all about. I am not worried about the ones who are coping - but
how do you take someone is not demonstrating adequate Captaincy skills and
give
them guidance. First you have to know what you are looking for!
BTW it seems that ACs and the rest of the crew have a different tilt on
Captaincy. At what stage in going from the right to the left seat does this
change of attitude occur I wonder.