Thursday, May 6, 2010

Grading Philosophy

Does the final grade for a student depend entirely on the numbers, or is there a certain element of whether or not they learned the material intrinsic in the final analysis? This was a discussion I had with a colleague of mine the other day when talking about our classes, and her argument made a lot of sense to me. Her point was that if the students did not do well on a particular exam, but demonstrated knowledge of that exam's material on the cumulative final, she would use the "grade" on the final section to replace the exam grade. She did not tell the students this (to avoid a lack of effort on the midterms), but put it into practice.

It is certainly easier to hold up the numbers and say that a student earned an 'X' grade based on their performance, but I've been thinking whether or not a strict adherence to the numbers actually has any value in actually rewarding the students for learning. This is particularly relevant to me as I sit in front of my class taking their final this morning, because their second exam was a blood bath with an average of 45%. But, about 25% of the final is on the same material, which should give me an idea about what they learned from that portion of the class. Opportunity or dangerous precedent?

Obviously, this type of policy would favor the students in the class who, to this point, are not doing all that well. Is it fair to give them the opportunity to boost their grade when others have done so by performing when they had to? Honestly, I don't know. Like with just about everything relating to teaching, I'm figuring this out as I go. If our objective, however, is for the students to learn the material during the course then it stands to reason that they should be rewarded for 'getting it' even if it doesn't click until the end.

16 comments:

  1. I think the idea has a lot of merit. The way I tend to get around the issue you mentioned in my smaller classes (less than 100) is to create a discretionary category that isn't immediately obvious to them (the total number of chances for attendance/participation was less than the total number of points I awarded for that category -- I created an extra 15 point category where I gave more/less points for engagement and effort). It's not the same as what you're saying.

    I think what I would do is use the mean of, say exam 2 and the final, to replace exam 2, in a student's grade where they bombed exam 2 but displayed competence in that material in the final. That way they don't do as well as the student who "got it" back in exam 2, but they also don't completely suffer for struggling earlier in the semester.

    ReplyDelete
  2. I honestly do not believe it is fair for the professor teaching the course to do the grading. It is too tempting to give a student who has worked really hard, and who has learnt a lot (but is still underperforming) a higher grade than than one who has not put in the effort. Grades should be for performance not endeavor. Grading should also be blind.

    Standard should be set, and it should be the professors roles to help all students meet those standards, not just figure out the best of that particular batch.

    Instead it is all relative, and your grade is dependent on your peers. As a result, a degree from one university is not comparable to a degree from another.

    Curved grading does not encourage students to peer-tutor. If standards are rising a professor is forced to raise the bar, and vice versa.

    I agree with you that end performance is what should mater the most. Exams throughout the semester should be an aid to learning, and nothing more.

    Thanks for giving me another reason why I do not want to teach at university - the system is fundamentally flawed. I would go insane. Good luck.

    ReplyDelete
  3. Thanks for giving me another reason why I do not want to teach at university

    I like to do what I can for the students out there.

    ReplyDelete
  4. This kind of grading definitely has merit.

    I understand the concerns over fairness to the students who have worked hard all semester, but I don't think they really hold up to scrutiny. First of all, we're making an unreasonable assumption in saying decent grades so far = hard work. In lots of cases that is true, but what about the kids who are lucky enough to get it from the beginning without much difficulty? In that case, the student with the lower numbers would have actually worked harder, but most teachers wouldn't dream of penalizing the higher scorer for "getting it" too easily. Why would effort matter in one case but not the other?

    Secondly, grading based on effort, while it appeals to our sense of fairness, is mostly a bunch of crap for an assortment of reasons. 1. Effort is hard to measure in an objective way. 2. Grades are kind of a flawed system for measuring a student's achievements and abilities anyways, but in my mind, if the goal of courses is education, grades should attempt to be a measurement of learning. What gets measured gets done--if you don't reward learning, then it won't be your students' priority.
    3. This is totally cliche, but *Life Isn't Fair*. Different people will have to put different amounts of effort into different endeavors to achieve satisfactory results. Factoring in effort with children in order to teach that hard work is valuable might be reasonable, but by college, it is your students' jobs to figure out how hard they need to work and just do it.

    I know tons of people that I have taken courses with that WORKED THEIR ASSES OFF in every science course they took in college and still got B's and C's... because they really hadn't mastered the material. It isn't fair that many times, I put fewer hours than those students and got my A's and B's. But there isn't a way to correct for that without making grades as they currently work practically meaningless... and even if there was, would we want to?

    ReplyDelete
  5. I'm not advocating changing a grade based on effort, perceived or real. All I am asking is whether a student's ability to assimilate knowledge should be looked at as "a whole", or in discrete periods that end in an exam during the semester.

    ReplyDelete
  6. "...whether a student's ability to assimilate knowledge should be looked at as "a whole", or in discrete periods that end in an exam during the semester. "

    I would argue that this specific question is very easy to answer:

    A: Grades are given on a per course basis;
    B: Courses are organized around a set of topics and a set time frame;

    therefore,

    C: Grades should reflect how well students master the material over the entire course period.

    ReplyDelete
  7. I once took a class where the exams prior to the End term exam only gave points to the end term exam... so that if you were good at the first two exams, you got added point bonuses to be able to pass/ get good grade on the final exam. That way you could be missing out on the first two exams and still end up good if that was the way you learned best. Or you had good grades on the first two ones and had less stress over the last exam.

    I think there is merit not to count the first exams as much as the "comprehensive big one" in the end. I guess I might be biased since I am better at learning over a longer time, after the labs and the lectures than to read the first two/three weeks ;)

    I do think however, whichever you decide to do, to put it in a category or a clear system since someone will inevitable ask you or complain about their grades and want to know "how can you do this to me" and then it is better to have thought it up before hand... Good luck! And I think you are to be applauded for trying some flexibililty rather than just adding % to % and getting a final grade. (but there is a reason one mostly ends up with that since it is easy to explain to the students...)

    ReplyDelete
  8. I think some of this depends on how much "flexibility" you can get away with as opposed to having a published, formal X%= grade Y. Here we have a fair amount of flexibility. My general policy is that I guarantee that certain % will be at least a certain grade, but I reserve the right to drop the lower limit of the grade range. So I might guarantee that 90-93% = A-. but some years I'll give out A-'s to people who make as low as an 87%. I also announce that I take trends into account for people near grade margins. So someone who bombed midterm1, but did well on the other midterms and the comprehensive final might get a half or full grade "bump" depending on how dramatic the effect was.

    The other reason I like a comprehensive final is it allows me to have a policy that students can skip one midterm for any reason at all, and count that section of the final as their midterm grade instead. I don't care if you are drunk, hungover, sick, on vacation, taking the MCAT,dead grandmother... I just don't give makeups. This helps good students who are genuinely sick etc, but students who skip a midterm because they aren't caught up or were hungover tend not to perform well on the final either. I don't have to judge "how good" excuses are, I just point at the policy. I do have to warn them, however, that if they simply blow off the first midterm and then are actually sick for the second, they are on their own...

    ReplyDelete
  9. I think you have to look at it as a whole. The fact of the matter is that their are students who "get it" but just are not good in timed, test type conditions. Which is why providing different types of assessment is ideal (ie presentations, papers etc). Midterms tend to be a crap shoot of which exam is coming up first, last etc. It never mattered how much I understood the material, I always did poorly on the last midterm from my group of classes. I just too tired and burnt out

    ReplyDelete
  10. If a student 'gets it' in a mid-term they presumably will 'get' the same material in the final exam. Averaging the two scores should make no difference. If a student finds it tough going to start with and gets a low score, but then puts in a lot of hard work so that they understand the material in time for a final exam then they would benefit from this procedure.

    Thus I think that, done right, the method can benefit both the naturally gifted and those who have put in a lot of work.

    ReplyDelete
  11. To me, bumping up grades is fine, but ONLY if it does not cause students with lower scores to get better grades than students with higher ones. Otherwise, it is really unfair to the students who are not favored. "Bumping up" grades is also extremely prone to being influenced by opinions about particular students. In my experience, even professors whose judgments I generally respect have a hard time not being influenced by how much they like particular students.

    "Whyme?"'s approach definitely seems the fairest way to go.

    ReplyDelete
  12. I've actually considered doing something like this in my classes, but the classes I'm teaching right now aren't well suited for it (grade mostly based on project write-ups, no exams). I think the bottom line should be if the student grasps the info/concepts by the end of the class, so adjusting grades upwards for students who do well on the final is totally reasonable to me. Plus, something like this gives the students that are struggling incentive to keep trying to improve instead of just getting frustrated with their earlier "failures".

    I don't think that this sort of approach has anything to do with evaluating effort per se, as some of the comments seem to imply. There's no random bumping, just a set of protocols that allow the students to improve their performances. And, so long as those protocols are applied to all students, there's no chance for cries of favoritism.

    ReplyDelete
  13. im not a big fan of changing the rules at the very end. i think it's good to be transparent in how you grade--not only will your students appreciate it, but if something you are doing is suboptimal it gives people a chance to give feedback.

    ReplyDelete
  14. This is a great question that I think illuminates one of the great ironies of teaching in Higher Ed. While we might be experts in our content areas, most of us are just thrown into the classroom as new faculty members with little idea of how to teach the content, let alone how to assess whether and how the students are learning the content. As it happens I teach Educational Assessment at the Undergraduate and Graduate levels. What we know to be 'best practices' in assessment is that products generated later in the grading term (semester in this case) should be weighted heavier than products generated earlier in the term. This would include labs, homework, exams, etc. Why? Because students come into your classes with various levels of experience and knowledge. What you assess early in the semester is more a reflection of what they already possessed rather than what they consumed in the course. Some students had stronger foundations on which to build the new knowledge. For others with weaker foundations, it might take them a little longer to firm up their knowledge base before they are able to master the new content. If the grade for the course is to reflect their mastery of course content, does it really matter if the student's learning curve was linear or curvilinear? I would encourage you to read a great article that appeared in the September 14, 2009 online issue of the Chronicle of Higher Education by Douglas Reeves entitled, "Remaking the Grade, from A to D." His is a really good comentary on assessment in higher education.

    ReplyDelete
  15. I think that echoes my point above - not all students learn the material when we expect them too. In my mind, they shouldn't be hugely penalized for not understanding some of the material early in the course, as long as they 'get it' by the end.

    The problem I see with telling students this policy is that the mid-term grade will not be good indicators of where they are at with the material because they will know there is a safety net. If I had been teaching the course for a decade and knew what the students generally understand at different time points, I might be more likely to do this. However, I need the mid-terms as much as they do to figure out what concepts I need to talk about again and which they have picked up.

    ReplyDelete
  16. I've been using a similar method for a couple of decades. Giving a student a B who has achieved an A mark on the final -- the same grade as the student who has plunked along at a B all semester seemed silly to me. My grade reflects what you show me you know at the end of the course. I do point out that blowing off the midterms is not a route to getting an awesome grade on the final, trying to run a marathon the first time out is not generally going to work. At my institution this doesn't seem to lead to students blowing off the earlier work.

    Ask yourself this question -- who do you want in your lab, the person who gets it quickly the first time, but doesn't put in a lot of work to do that, or the one who, seeing his C+ on the midterm, digs in and gets an A on the final?

    ReplyDelete