In very much the same way, in many applied scenarios, while we may care medically, what matters is that final output, that final piece, but it is hard to track back without doing assessments and careful countermeasures.
DOUGLAS WILMORE: I have sort of a broad generic question for the group of people who talked about behavioral things. It would seem to me that it would be worthwhile trying to think about how you could build a monitoring system into the actual task that the person was doing. I think Mary tipped me to this idea. If you had a computer operator, you could build the ability to shift or to spell wrong or even time variables into that particular task. And Harris, if you are looking at these guys shoving shells into that big gun that you had, it would seem to me that you could time throughout the day the speed with which they were loading the big gun or something like that.
I was wondering if that is not the evolution of all of this, that instead of asking people to stop what they are doing and do their cognitive function testing, to somehow build a monitoring system into their specific task. It may even get to the place where that foot device that looked at how many steps people take could even be programmed to do that, because you slow down and you get sore feet, and all that sort of thing.
HARRIS LIEBERMAN: Some of those things have actually been implemented, in part. Certain occupations that require the use of a computer to do a job are monitored by the supervisor or a higher level to measure individuals' efficiency on a day-to-day basis, and they are given feedback or they can even be fired if their performance does not meet the specs. What they will do is they will look at the average specs of a telephone operator because that is all automated right now, and if somebody is falling off the bottom of the distribution, that person is in significant trouble. There are some ethical questions there, certainly, but that can be done legally right now.
In terms of actually what you are asking, what is their final output, are they firing more shells? It turns out that in that particular example it is extremely difficult to get enough information to relate it back to any measure of performance, because there is so much variability in the individual unit. Some batteries are just better than other batteries, so when you try to get the individual performance data, it all gets confounded with whatever treatment you have got out there. Now, in a situation where, say, there were environmental variations over the course of a long period of time, it might be possible to relate that to some kinds of performance in a correlative kind of way and that would be valuable. That is a good idea.
MARY MAYS: Another really good example that is relatively easy to track is the motion of your eyes, where they are going and what they are looking at and so