Rob Weisskirch wrote: > > Our university has jumped on the assessment bandwagon and those who > have drunk the kool-aid talk about "assessment of student learning" > and looking at student evidence. I continue to ask why looking at > grade distribution is not an indicator of learning.
Rob, Although I'm chiming in a little late on this one, I think that a little statistical test theory explains what is going on with crystalline clarity. First, allow me to introduce two technical terms: The "*sensitivity*" of a test is the degree to which is correctly identifies members of the category that it is intended to test for. (In a school grading system, the proportion of those students doing "excellent" work who get an "A".) By contrast, the "*specificity*" of a test is the degree to which it correctly identifies non-members of the category that it is intended to test for. (In a school grading system, the proportion of students doing non-excellent work who get a non-A.) Tuition-paying parents decided a while ago that our grades weren't sensitive enough (i.e., that we were giving too many of their "excellent" Johnnys and Janes non-A grades). Our administrations (and we) gradually caved in to the pressure (fearing law-suits, non-contributing alumni, "hard" reputation, etc.). Because we could think of no better way to substantially improve our grading procedures (increasing both their sensitivity and specificity), we simply moved the goal posts, giving more student higher grades, and reducing failure to nearly zero. This probably increased the sensitivity slightly, but at the cost of massive decreasing the specificity. This process is called, in the vernacular, *grade inflation*. After a while, those very same tuition-paying parents, now in their alternative capacity as employers (and voters, don't forget), began to realize that lots of job applicants were coming at them with degrees and high grade averages, but without the actual skills that were once expected to accompany such degrees and grades. Now these tuition-paying-parent-employer-voters declared, in outrage (at, note, the very result of their own earlier demands), that our grades weren't /specific/ enough (i.e., too many (of other people's) "non-excellent" Johnnys and Janes were getting A grades). Since the grading system we were using (at their earlier insistence) didn't adequately distinguish between "excellent" and "non-excellent" students, they declared that we were incompetent, and that the government (or its proxy) should jump in and and see how well students learn specific things without, note!, those assessments being attached to particular students (thereby damaging their precious self-esteem and job prospects) but, rather, to their schools and teachers. In this way, they will be able to have the best of both worlds. The grades for *individual* students will remain highly "sensitive" (because many students will continue to get high grades), but the assessments done of the institutions and teachers will be highly "specific" (teachers whose student do not meet externally-developed criteria will take the bulk of the blame). See, it's easy! :-( Chris -- Christopher D. Green Department of Psychology York University Toronto, ON M3J 1P3 Canada 416-736-2100 ex. 66164 chri...@yorku.ca http://www.yorku.ca/christo/ ========================== --- To make changes to your subscription contact: Bill Southerly (bsouthe...@frostburg.edu)