Thanks Keith Hendy for your thoughts below! You identified for me what was
niggling in the back of my mind about the discussion on rogues. Any
attempt to characterize individuals needs to be supported by rigorous and
valid research. It would seem to me that Major Kern's work is based on a
large sample of case studies and thus may be valid. However, many of the
definitions which followed seemed to be largely anecdotal and thus can only
be interpreted when added to a significant sample of such data. However,
you are also quite right in saying that such a classification needs a
criterion to be measured against and it is highly unlikely that "rogues" or
any other "type" will be likely to have more accidents than any other.
(Yes, as far as I know, the search for "accident proneness" as a
personality trait has long since been abandoned).
I would suggest that attitude and personality are more likely to, knowingly
or unknowingly, have and effect through the influence they exert on others
in the organization. Consider Reason's (1991) hypothesized list of
"violation producing conditions" (that is precursors to individuals
violating rules):
1) lack of organizational safety culture;
2) conflict between management and staff;
3) poor morale;
4) poor supervision and checking;
5) group norms condoning violations;
6) misperceptions of hazards;
7) perceived lack of management care and concern;
8) little elan or pride in work;
9) macho culture which encourages risk taking;
10) beliefs that bad outcomes won't happen;
11) low self esteem;
12) learned helplessness;
13) perceived licence to bend rules;
14) ambiguous or apparently meaningless rules, and;
15) age and sex (young men).
How many of these conditions are likely to be precipitated or exacerbated
by the presence of the character type identified by Major Kern,
particularly if these individuals are in senior or training positions?
Thus, it is through their effect on safety culture which "rogues" (if they
do exist) will do most harm.
Having said that, I don't know what to do about it. It seems unlikely that
one would be able to define the characteristic enough to select individuals
out. By definition, "rogues" themselves are probably resistant to training
in particular and change in general.
Just some thoughts,
Joel.
At 12:59 19/08/97 U, you wrote:
> RE>>Roguism 8/19/97
>
>As I follow this thread I am reminded of the somewhat shaky literature on
>'accident proneness'. The last time I checked, I believe that attempts to
>define the cadre of more accident prone drivers had fallen on barren ground.
>As appealing as this concept is (and of course we all KNOW the profile of the
>accident prone: drive red cars ooops that's what I drive, little old men in
>hats driving Chev Wagons, women, various ethic backgrounds - substitute your
>biases here, BTW I am NOT stating MY opinions in the former list of
>characteristics) there may not be good scientific grounds for promoting this
>idea. Certain factors do have predictive power: age/experience (correlates
>highly with proficiency), use of alcohol etc. I am not sure that attitudinal
>and personality factors have been as successful.
>
>Now if the so called rogue is not more accident prone that the non-rogues
then
>we just have to write him/her off as a pain in the butt. If the
>characteristics of the rogue do make then more likely to have an accident,
>then obviously we should try to do something about them.
>
>So the question I end with is, can the so called rogue really be identified
>with an acceptably low rate of false positives? Is there anything other than
>observational/anecdotal data to support this position?
>
>
>Cheers
>
>
>Keith Hendy
>
>DCIEM
>
F.J. Joel Morley
Research Officer
Human Factors Technology Group
College of Aeronautics
Cranfield University