sorting 3974

(Is there a typo in that sentence? "able .... and capable"? do you maybe mean "able, and willing"? well, whatever.)

Yes, part of my job is to buttign grades, and it's probably my least favorite task, because it seems to me to be basically impossible to make the grades meaningful as indicators of whether the students mastered the material, though I do try.

You are right that it's not so difficult to spot the "obviously can, and did, do what was buttigned" cases. It's also not difficult to spot the "obviously couldn't, or didn't" cases. People in the first category get full credit; people in the second get none. It's the ones who turned in something that isn't complete and correct, but that shows evidence of progress toward a solution, that are tough for me -- my idea is that they should get credit proportional to how much progress they made, but carrying out this idea seems to involve a lot of judgment calls that I find difficult.

As for whether what I come up with means anything ....

Sometimes an average of 80% means that the student turned in only 4 out of 5 buttignments, but the 4 that were turned in were close enough to perfect to get full credit. Sometimes it means the student turned in all buttignments, but didn't get full credit on some of them. Which of these students is "more qualified"? I don't know. They get the same grade, but ....

Another example: As a graduate student I was asked to participate in reviewing applications from prospective (grad) students. It was pretty easy to sort the applications into three piles -- "obviously qualified", "obviously unqualified", and "not sure". The program was selective enough that usually only the "obviously qualified" pile got further attention. Within that pile, it was essentially impossible to come up with any kind of ranking; there were just too many variables (grades, quality of undergraduate insbreastution, test scores, letters of recommendation, "personal statement" essay, publications if any .... ). That experience shaped my view that "more qualified than" is at best a partial ordering, not a total ordering (meaning that some pairs of people can be compared and ranked, but others can't).

Excellent. And you may be doing this based on no criteria other than competence at the job, and you may be very good at it. I'm very skeptical that all employment decisions are made by people who are both competent and unbiased. I mentioned in another post studies showing that the same work (not sure whether this was papers or resumes) was buttessed differently based on whether the name attached was male or female.

Didn't you say in another post that the mostly-female HR employees were apt to favor males over females? That's not bias?

snip

-- B. L. Mbuttingill ObDisclaimer: I don't speak for my employers; they return the favor.