Monday, November 07, 2016


Rigor and Fairness

I’ve read that the hallmark of a great mind is the ability to hold contradictory ideas simultaneously.  By that definition, academia is full of great minds.  Many of us manage to believe both of the following:

  1. Grading is a core component of academic freedom, and should be left to the professional judgment of individual faculty barring extraordinary circumstances (i.e. harassment).
  2. Students’ performance should be judged consistently across sections, and their GPA’s should not be held hostage to which section they take.

When you have, say, twenty people teaching Intro to Field, and one of them is a conspicuously, significantly, and consistently harder grader than the rest, to which belief should you default?

I thought about that in reading the IHE story about the adjunct who says he was fired for maintaining academic rigor in his classes.  Maybe he was, maybe he wasn’t, but I’ve heard enough variations on that line over the years to want to offer some possible context.

Over the years at several different colleges, I’ve seen faculty who say that the reason students don’t like them is that they have high standards.  That sounds reasonable until you see other professors with high standards whom students love.  Having observed my share of classes over the years, I can attest that every so often, rigor can be the last refuge of the misanthrope.  

One clue is polarized feedback.  If student feedback on Professor Scrooge is evenly split between “best ever!” and “run away!,” that’s consistent with rigor.  If student feedback is uniformly negative, there’s likely a reason.

That said, I have to admit that yes, sometimes administrators will use pass rates to judge faculty.  At DeVry, I came under pressure to do that, which was a major reason I left.  In the very short term, you can increase pass rates just by passing anybody who can fog a mirror.  But in the long term, that approach is both unethical and institutionally suicidal.  I’ve been consistent in avoiding that and in instructing my deans to avoid that, precisely because I’ve seen what happens when standards fall.

In the context of a community college with a strong tradition of transfer, too much laxity on grading will lead to underperformance by students at destination colleges.  If too many of our students crash and burn at destination schools, the destination schools would eventually either stop taking them entirely or make them jump through so many hoops that they might as well start over.  

To me, the more elegant and sustainable way to handle ideas one and two is to go around them.  That could mean departmental norming sessions for grading papers.  It could mean trading papers and grading them blind.  It could mean agreed-upon rubrics.  It could even mean longitudinal tracking -- do the students who get A’s in Intro to Field do well in Field II, Son of Field?  If they don’t, that’s a red flag.  The red flag could signify any number of things, but it’s there.  

From this side of the desk, I’ve heard far more students complain about opaque or illegible expectations than about high ones.  For the most part -- and yes, there are exceptions -- students respect high standards when they understand them.   I remember in my own student days, the most frustrated I ever got with a professor was when a weak paper got an A-, and a subsequent stronger paper got a B-.  The overall level was probably about right, but the apparent disconnect between what I produced and the grades I received meant that I had absolutely no idea what was valued.  I had received tough grades before, but they made sense; I knew when I was struggling.  In that class, they didn’t make sense one way or the other.  That was maddening.  (To make matters worse, the professor was consistently inaccessible, so I couldn’t even ask the question.)

I’m happy to back up high standards when they’re clear.  If Professor Scrooge expects students in Field III, The Field Strikes Back to go above and beyond, so be it; college is supposed to be hard.  But if Professor Scrooge is miles away from everyone else in the Field department, it seems fair at least to raise the question.  Starting with threats would be counterproductive, but maybe a peer observation or a norming session would be in order.  

Wise and worldly readers, if Professor Scrooge is dramatically and consistently tougher on students than anyone else in the department, what would you do?  For extra credit, try not to violate ideas one and two...

If I were an administrator, I would be very, very careful not to impose my grading standards on anyone else. Generally the first thing to do is to let all the faculty in the department have access to the grade distributions (not individual grades) of all the sections—perhaps over several years. If the faculty all saw that one of their members was way out of line with the rest (either much easier or much harder, both of which are equally problematic), then the department faculty could start internal discussions about what to do.

They might decide that the old-timer who has not bowed to grade inflation really is worth supporting, and that the rest of them have been giving out too many easy A's (a common problem in "elite" schools). Or they might pressure an overly harsh grader to use the same standards that they do.

If there is a series of courses with some prerequisite to others, then looking at how good a predictor the grade in the prerequisite is to the grade in the subsequent class (both controlling for instructor and not) can let the faculty know whether any of their grading is predictive.

The idea of swapping grading only works if the different sections really are teaching exactly the same thing (which may be common in community colleges, but is very rare at the university level—if classes are teaching identical things, they are merged).
re: multiple sections at the university: Morning calculus is probably pretty similar to afternoon calculus.

The real issue with swapping grading -- at least in math -- is that small changes in the exam can result is large differences in difficulty. Testing implicit differentiation by asking students to find the equation of the tangent line to sin(xy) = x at a specific point? If I want the exam to be hard, I pick (1, pi/2). Easier exam? The origin.

Unless you want to set common exams and have group grading. Ugh. Good luck with that. My thoughts and prayers are with the Calculus Czar.
I thought the warning flag in the Colorado story was this: "Set aside five class sessions for helping students with writing skills and learning how to write an essay." That suggests a problem in the composition pre-requisite for that philosophy class or, if it doesn't have one, their entire curriculum. Do the universities that are to accept this course have two weeks of writing remediation built into their schedule for their version of that course?

Regarding having "Scrooge" in course three, I have wonderful memories of some bad and good choices in assigning faculty to a sequence. (This was a large lecture with the same person doing multiple lectures with common exam level and grading policy across the entire block.) Bad: easy person the first term followed by toughest and most misanthropic person in the department in the second term. Awesome: tough person in the first term with the easy person in the third. This is in a sequence where the difficulty level is naturally higher in the second term and lowest in the third.
I thought the warning flag in the Colorado story was this: "Set aside five class sessions for helping students with writing skills and learning how to write an essay." That suggests a problem in the composition pre-requisite for that philosophy class or, if it doesn't have one, their entire curriculum. Do the universities that are to accept this course have two weeks of writing remediation built into their schedule for their version of that course?

Huh, that was the only one of the recommendations that didn't strike me as an immediate red flag. Specifying the exact number of class sessions is a little too micro-managery for my tastes, but the basic principle (teach the students HOW to meet a higher standard) is fine. Ordering the professor to lower the standard is not fine; it defrauds the students who actually want to learn and have a reasonable expectation that the courses they take at the community college will be the equivalents of those at the university. It's also unfair to the students who pass the course and then discover they're in over their head when they take a higher-level course after transferring, and to the faculty who have to teach those higher-level courses. (Granted, since most students in this case are probably taking this as their one and only philosophy course to meet a gen ed requirement, this is less of an issue than it would be in a course where most students are expected to take a higher-level course in the same subject -- but still, the students have a right to a class that will set them up to succeed if they choose to do so.)
One way to get around this is to teach across sections so that one person teaches the same content in every course but several teachers teach one course.

Each lecturer sets 2 or 3 questions for the final exam (out of 6) and marks them.

It's not going to work once the number of section starts heading towards 4 or more.

[At my university, in the old days, the lecturers would set the exam and then the completed scripts would be sent by boat to Cambridge (IIRC, otherwise Oxford) to be marked and returned by boat. Graduation was 6 months after the final exams in order to have time for the turnaround. That's stopped now and the only hangover is that graduation is so long after exams.]
What is the purpose of grading? Do we want to signal whether students have mastered the material? Do we want to use it as a student ranking device? Should it be used to discourage students from continuing in Field X?

With respect to standards, how high is too high and how low is too low? If every student in a section gets an A or B, is that okay? What if a professor only hands out one A and 3 Bs out of five sections?

In an old job, I saw the paper Scantron forms that professors turned in and had the opportunity to do reports and queries to compare professors and sections over time. Some professors would use the side of a pencil to shade the A column all the way down the page. Others had such high standards that they had a 50%+ drop rate in their classes and typically awarded less than one A per section. One professor consistently had over 75% of his students drop or fail.

I think #2 is far more important than #1; it's a mistake to treat standards primarily as a matter of "academic freedom" for individual professors. Some professors are jerks. Some professors have standards that are so low they barely exist while others are loath to allow more than a couple of students per section to earn a B. Standards should be worked out collectively to reflect the goals of the institution. Faculty should be shooting for the same target even if they sometimes disagree whether a student has hit the mark.

Different institutions may have different goals. If Harvard or MIT wants to use Honors Calculus as a weed-out course for their math majors, that's fine. But Local Community College probably just wants to ensure that Calc I students will be able to do well in Calc II or to succeed at Regional University. If Lone Professor at Local Community College decides that his students should be held to the Harvard/MIT standard, he's just being a tyrant.
I've taught at an institution (1980-1987) with common exams both in intro accounting and in first year calculus. (Mid-sized state school--about 12,000 undergrads when I was there--in Illinois.) I don't know if they still do it. It required scheduling so that all the students took the exam at the same time (evenings) and a lot of work to get the exams written, because everyone teaching the courses wanted input on the content of the tests. (I frequently had students taking the accounting tests who had to miss my class--I taught a lot of evenings--and occasionally I had to make arrangements for a student to take my tests at a different time.) My understanding, from talking to the accounting faculty, is that it worked OK for them/
Fretful Porpentine @7:20 AM

But one of the other changes, cutting content by 20%, was necessitated to make room for 5 days of remediation for the shortcomings of their composition classes. I would also infer that limiting papers to just a few pages has been done for the same reason. My college has composition classes that teach students how to meet the higher standard of college-level writing for subsequent classes that require major papers, and uses prerequisites to enforce that.

Anonymous @5:32 AM

Calculus at different times of day can be as different as the people teaching it at my CC. Students who take calculus at a nearby university report vast differences, where one section has all of its exams taken on a computer using a commercial program while another uses hand-graded tests.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?