more on Brainwave fault explains slip-ups

[Dave Farber <dave () farber net>]

Delivered-To: dfarber+ () ux13 sp cs cmu edu
Date: Sat, 20 Dec 2003 12:14:37 -0600
From: Don Norman <norman () nngroup com>
Subject: RE: [IP] Brainwave fault explains slip-ups
To: dave () farber net
Geesh

Just after Dave Farber posts my contribution to RISKS, he also posts this
silly nonsense:

>
> From: Dewayne Hendricks <dewayne () warpspeed com> Brainwave
> fault explains slip-ups Scientists have found an explanation
> for those mornings where you put coffee on your cornflakes
> and the cat in the washing machine.

Stupidest thing I ever heard.  And I suspect it has to do with the
journalist, not the researchers (No blame to Dewayne, either -- he just
posted it. [AS DID I djf ]) .  The researchers being quoted, are from the 
Cambridge (UK) MRC
unit, one of the best units in the world, so I suspect their research is
solid, but those comments are either out of context or said by an academic
researcher with no understanding of the real world, but being forced to
answer some question about "what use is this research?"

The notion that P300 (a brain potential that occurs roughly 300 msec. after
a signal)  explains errors such as 3-mile Island or Chernobyl simply points
out people's desire for really simple explanations of complex phenomena.

> But the researchers, who presented their findings to the
> Physiological Society Conference, said the change could have
> more serious consequences.
>
> They say it may have been a factor in accidents like the
> Chernobyl disaster.

Bah and humbug.

Real world accidents are very complex. Invariably, there are multiple
factors, and had any one of them not happened, there would have been no
accident. At Chernobyl, among other things, the workers were testing the
reactor and in order to do the test, they disconnected many of the safety
features. That's only one of many problems. Another was that the reactor was
fundamentally unstable at low power levels, which is how they ran the test,
mind you, with safety features off.  Why were the safety features off?
Because otherwise they couldn't do the test. The number of failure points
here is huge. One sleepy operator or one bad decision was NOT the primary
cause of the incident, although these factors were also present (the test
was being was run in the middle of the night so as not to disrupt daytime
requirements)
.
On top of this, the simple neat trick of measuring brain potential makes it
sound like all accidents are caused by some silly stupid human, so if only
we could monitor people, we could tell when one of us was going to make an
error and slap our wrists or something.

What simple-minded drivel. Real world-accident investigations (when done
properly, and the National Transportation Safety Board (NTSB) reports serve
as excellent  examples of how to do them well), would never reach such a
conclusion.
Shame on all who propagated this nonsense.  More shame on the fact that now
this will only reinforce the notion that human error is the culprit behind
accidents, when it fact it is almost always poor design plus business
conditions and a profit-driven culture that force people to violate
procedures in order to get the work done. (How many of you paste the
password on the front of your monitor, or use the same password on multiple
accounts, or ... ?)

Don Norman
Nielsen Norman Group & Prof. Computer Science and Psychology
Northwestern University, norman () nngroup com http://www.jnd.org

-------------------------------------
You are subscribed as interesting-people () lists elistx com
To manage your subscription, go to
 http://v2.listbox.com/member/?listname=ip

Archives at: http://www.interesting-people.org/archives/interesting-people/