Let me add a second thought to Stuart's point about the reliability of
our measurement. I have concluded that my empirical test scores are not
distributed symmetrically around the "true" scores, instead they
likely underestimate the "true" score over the course of the semester.
My argument is based on the following observations. Over the course of
the semester, I see students commonly have a test score that is lower
than typical due to other issues--typically they have a cold or some
other medical condition that is not crippling enough to make them miss
the exam. For example, I had a student who had an 'A' average. He took
an exam while experiencing the onset of a migraine and made a 'B' on
that exam. Mathematically, this pulled his class average down to an A-
(by 1 point). But the A- didn't represent his general pattern of
performance. Instead it represented the mathematical effects of a
single clunker grade. So he received an 'A' because that represented
his general pattern of performance.
Ken
"Stuart McKelvie" <[EMAIL PROTECTED]> 08-May-07 4:14:01 PM >>>
Dear Tipsters,
The general question is this:
When we calculate a final percentage grade, how reliable is it? Another
way of putting this is: What is the standard error of measurement?
Sincerely,
Stuart
______________________________________________
--
---------------------------------------------------------------
Kenneth M. Steele, Ph.D. [EMAIL PROTECTED]
Department of Psychology http://www.psych.appstate.edu
Appalachian State University
Boone, NC 28608
USA
---------------------------------------------------------------
---
To make changes to your subscription go to:
http://acsun.frostburg.edu/cgi-bin/lyris.pl?enter=tips&text_mode=0&lang=english
