damn, boy!

Wow, you even outdid me. I was way too lenient in the grading.

Yeah, it was painful. I put too much weight on the homework average, and it was too late to do anything about it. ☹

Next semester… a whole other ballgame.

I used a curve based on the mean and actually went beyond what my syllabus required, which is how I ended up with inflated grades. I’m considering going with a proportional curve where every grade is: (your score / highest score) * 100. I also didn’t give any points for homework—only the inflated tests mattered—and that will change next semester.

The basic failing here, I guess, is that several people who did not deserve to pass, and who did not understand macroeconomics, received passing grades. That creates a good deal of shame on my part.

I’ll also be a little wiser next semester—I expected the students to actually READ the book before coming to class. That’s why homework will count some in the coming semester.

In my case, a proportional “curve” wouldn’t have helped much; I’d have had to be giving B’s for 95’s. I’d have had to calibrate the grades to a normal distribution or something, or resort to less transparent (i.e. much more arbitrary) assignment of letter grades.

On the other hand, I could solve the grade issue real easily: if I got rid of (mandatory) homework, a lot more students would do poorly on the exams. Personally, I’d rather they learn the material, even if the grades end up high…

Lots of little lessons learned from this semester; tomorrow’s big academic project may be cataloging them all.

Our cases are fundamentally different. I’m dealing with a wide range of students, many of whom couldn’t be bothered to either attend class or read the book. Part of the difficulty of teaching at a non-selective university. My students ranged from the very bright—maybe five of 60—to the, well, not so bright to the point of, well, not good.

My reasoning behind a proportional curve is to motivate the laggards. Everyone in the class can score in the 80s and still get either a high B or and A (I’m using the ten point scale) based on their proximity to the highest grade.

The tests I gave were intentionally quite difficult, to see if they were grasping the concepts. Most of the questions were oblique in their approach to the concepts we discussed. Some gimme questions were included—like the slope of a supply curve, and I still got answers saying it was horizontal—but most were difficult. That, combined with no mandatory homework, led to low grades. The high scores on the tests 89, 97 and 84, IIRC. The 97 on the second test was an outlier.

The class scores did seem to prove the central limit theorem, though. I didn’t check the mode, but the median and mean were identical in every case and the grades seemed to follow a normal distribution.

How did you grade the papers? I found in previous years of teaching methods that broad paper assignments with few parameters (to “give the students the maximum academic freedom”) forced my hand into more lenient, highly subjective grading schemas.

Our poi sci methods is designated as an “intensive writing” course which requires…at the very minimum….a 15 to 20 page research paper that has gone through at least one re-write/ graded rough draft. I have found that by laying out more parameters, I end up being able to grade the first drafts with a strong hand. The students don’t get in an uproar because they know their paper grade can be improved with the final draft and the rough draft grades keep inflation to a minimum.

Actually, I found a loophole in the language for our “intensive writing” rules so the last time I taught methods, I had them write two smaller papers (for one of the papers, they could bring me a rough draft for ungraded comment if they wished….thus meeting the letter of the law with respect to rough drafts). This allowed me to assign one highly structured paper (which could be rigorously graded in a fairly objective manner) and one less structured paper. This scheme turned out to be easier on me as far as grading went and the rigorously graded paper kept grade inflation down.

Another thing I did to keep the homework grades from inflating the final grade was to require some of the “home“work to be done in class as a group (group size usually 4 people). The logic was: they learn a concept by (1) reading about it (2) hearing about it from me and seeing me work examples on the board and (3) “teaching” it to each other in groups. It works pretty well and has the side benefit of keeping homework averages down for those who skip classes (you can’t do the group “home“work if you aren’t there to be in a group)

Finally, I DO NOT give open book/ open note tests. I DO give a formula sheet, but the formulae are not labeled and they have to know when to use which one. My final exam scores, which tend to cover the same topics you covered in your last final, tend to be slightly lower than mid-term exam grades.

On the paper grades, they ranged from 21–25 out of 25, but then again there were no genuinely bad papers.

I’ve done the in class group “homework” thing, but didn’t grade them.

And the final exam grades actually weren’t a problem, inflation-wise. It was the nearly-universal 100% on homework and the good midterm averages that really blew the grades up.

Working on the syllabus today, I went with:

10% homework (down from 25%, probably still guaranteed 100% if you do them)
10% group project (100% if you help, 0% otherwise)
20% paper (down from 25%)
30% midterm (up from 25%)
30% final (up from 25%)

I think that will fix things.