but, this is why we don't use one item on a test ... we use not only
multiple items on tests but, multiple tests ...
this is why we don't use one course in college to decide whether we
graduate a student or not ... we use 40/50 courses ...
that's why we don't let someone graduate (in theory of course!) from high
school based on their work in the third grade ... but, we have a cumulative
record over a dozen or so years ...
unfortunately, it is true that too many decisions are made based on ONE
thing ... and we have to shake this bad habit ... but, that does not change
the notion that at THAT point in time ... the measure might produce data
that are highly reliable for making inferences about that trait ... at that
moment or a parallel moment in time ...
the problem of course is that we take the one instance and then over
generalize ... THAT is where we err ... BIGtime ...
At 06:28 AM 9/13/00 -0400, Alan Zaslavsky wrote:
>The following article may be of interest to some of you who are trying to
>get across the notion of reliability, particularly those who are teaching
>H.S. or young college students who have recently gone through high-stakes
>achievement/competency testing programs. You can also download directly
>from the New York Times web site at
>
> http://www.nytimes.com/2000/09/13/national/13LESS.html
>
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
