Originally posted by Symmetric Chaos
I don't know, you might be getting a skewed perspective from psych and neuro professors who probably more aware of controversies in test construction and operationalization.
maybe, though, I took nearly as many sociology/anthropology courses in my undergrad as I did psych/neuro (I had the prerequisites to get a minor in middle east studies), so I am certainly aware of how testing and grading is done in those courses.
A huge issue in most cases is labor. In highschool and low level university/college courses, the work load that asking students to write papers or do more challenging things other than multiple choice produces is insurmountable. I taught a section of into-psych that had 300 students in it. That was one of 12 sections. To test the students on issues of grand comprehension or cognitive maturity with the material would have taken an army of markers that the university wouldn't pay for, and probably wouldn't be available there if they would.
Additionally, and I could rant on this, there is a lowering standard for students (almost ever prof I have talked to has mentioned this) and the business-ification of universities has made core courses much more about giving the students the credits, thus taking more of their money and getting them degrees, than about making sure they have an actual comprehension of the material. For instance, in the intro-stats courses I was teaching this past year, the assignments were designed such that students could only lose a certain number of marks per question, and in many cases even if an assignment was left blank a student wouldn't lose enough marks to fail. This is because intro-stats is a core course for a psych degree, and thus designed by a committee of individuals whose primary concern is, "how can the most number of students pass so we can keep taking their money?"