• About Me

    I’m an opinionated Grumpy Old Man. I enjoy the intellectual give and take that goes along with that, but have very little patience for stupid people (Note: there is a big difference between “stupid” and “educated”. Some of the stupidest people I’ve ever met have a PhD…). Beside arguing, I like to build things in almost any media. Right now I’m mostly building in wood, Lego, and a bunch of different electronic media. I teach in a number of different venues - from preschool all the way through graduate school. Subjects range from talmud to neuroscience to engineering.

    For fun, I like to bash people with swords (OK, so they’re made of foam. It’s still fun). Although I spend a lot of my time in a wheelchair, I manage to keep pretty active (Like bashing people with swords). I am a libertarian, and have a hard time finding anything good to say about government or politicians. OK, politicians might make good sausage, but that's about as good as it gets.


  • VOTE!

    The question: Sould I use more graphics in my posts? I'm lazy, so I don't use them too o ften. If you think more pics would make the blog better, let me know! Simply let me know - email sphyrnatudevote@gmail.com with your vote!
  • Ask Dr. Science has MOVED

    Dr. Science now has his own Blog, so cruise on over to: http://askdoctorscience.wordpress.com to see what's cooking in the lab!
  • June 2007
    M T W T F S S
    « May   Jul »
  • Meta

  • Advertisements

Is there a better way to grade students?

This is the culmination of a bunch of different discussion, blog threads, and years of using and getting grades. I’m talking about the old-fashioned A B C D F +- grading system that most of us grew up with, and which many of us still deal with today.

When I was kid, a letter grade was supposed to be a reflection of how much of the given material in a class you had mastered (or at least mastered well enough to take the test, but that’s a different subject). If you learned everything, you got an A. If you learned less than 60% of the material, you got an “F”. The breakdown was pretty easy to understand, and in all of the many school systems I attended, were pretty much the same:

97-100 = A+,  93-96 = A,  90-93 = A-

The 80’s were worth a “B”, 70’s a “C”, 60’s a “D”, and anything less an “F”.

Simple, effective, and everyone knew what it meant. Of course, if you moved from school system to school system as I did, you quickly learned that the material in classes with the same names can be completely different. The “good” school systems can cram an awful lot of material into a class, but the same class taught in a crappy school will have maybe 1/4 of the content. Of course, the grades are the same (more on that later)…..

So what is the purpose of the grade anyway? From my point of view, there are a couple: first and foremost, a grade should indicate the level of mastery that a student has achieved in a given class or subject. If it does this, a grade will automatically meet the second requirement, which is to communicate this information quickly and efficiently to other people, who may or may not know the student. These people could be parents, staff at a school the student is transferring to, the admissions group at a college the student is applying to, an employer, or anyone else that cares about (and has the right to access) this information.

The problem with letter grades is that they reduce students to a percentage. If they are used properly, all the letters mean is that the student mastered X% of the course material. It doesn’t tell anything about how much the student learned – (s)he may have known everything taught in the class before ever entering that particular classroom. But is this a problem? If I am looking at a students transcript (either on a University admissions board, or as a potential employer), and the transcript shows that he student got an “A” in English Composition, Creative Writing, and Masterpieces of Western Literature, but got Cs and Ds in Chemistry, Biology, Algebra and Geometry, I have a pretty good idea that I don’t want to place that student into a technical job or college program. On the other hand, if I’m looking for someone to write manuals, fill a slot in an English Department, or some such, this candidate may be OK. The letter grade has made it very easy for me to perform an initial sorting of this candidate, and make an initial determination as to his/her appropriateness for whatever slot I am looking to fill. I don’t care where the student learned the material, what I care about is that the student actualy knows it. If the student took a class knowing full well that (s)he already knows all of the material, so what? They get an easy “A”, and their transcripts reflect the fact that they have mastered that set of skills.

Similarly, if a student entered a class with none of the prerequisites, works his/her butt off, and actually manages to catch up a bit – learning a lot of material in the process – what I want the grade to reflect is the portion of the course material that the student mastered. If Joe takes calculus, but doesn’t know how to add, subtract or multiply, and by the end of the semester has mastered basic math, algebra, and geometry, that’s great for Joe, but he didn’t learn any calculus, which is what his grade in the calculus class should reflect. In other words, an “F”. Is this fair? Joe has definitely demonstrated the ability to master mathematics, the dedication needed to learn a lot of material in a short time, and the willingness to stick to it and get it done. Shouldn’t his grade reflect some of that effort, diligence and ability? Absolutely not. His grade in calculus should be a reflection of his performance of calculus. If he wanted to get a good grade, he should have taken the remedial math, algebra and geometry course, then taken calculus. Given his performance, it is extremely likely that he would do well in calculus – if he come to the class with the proper basic skills. If we feel sympathy for Joe, and give him an inflated grade as a reward for his effort and progress, all we are doing is undermining the usefulness of grades. It doesn’t matter how you report a grade in a case like that, unless you give Joe his “F”, you are simply making whatever grading method you use meaningless.

Here’s another example: You are teaching Spanish, and a student comes to you and tells you that they are doing poorly in Spanish because they skipped the pre-requisite to chemistry and have to work really hard to pass chemistry. They are doing well in Chemistry, but in order to catch up in Chem, they have had to cut back on their Spanish studying. Is this grounds to soften up your standards and give them a better grade than they have earned? Of course not (at least I hope that was your reaction).

The fact that different schools have different curriculum and grading standards (good vs. bad schools) does cloud the water a bit, but if I’m sitting on an admissions board, the board is aware that an “A” from this crappy school is the same as a “C” from that good school, and believe me, we take that into account. If I am an employer and I’m looking at recent graduates (about the only time academic transcripts really make a difference), I’m either fairly familiar with the Universities and colleges that have programs that provide my employees, and know what their grading (and probably their curriculum) means, and adjust accordingly. (A least that was the case when I was hiring.) In the event an Admissions group or employer wants more detail about the students ability, an interview, essay, or any other more personal follow-up can easily provide much more information than something a teacher wrote.

So where do letter grades really fail? Well, the students that simply cannot take tests will tend to get bad grades – at least in situations where tests determine the grades. Is this a problem with using grades, or a problem with how the grades are generated? An unfortunate reality is that most of our schools are test centric, and programs like “No Child Left Behind” only exacerbate the problem. The students that do not test well do get poor grades. However, rather than blaming the reporting mechanism, blame the evaluation that generates the report. No matter what grading system you use, a faulty evaluation will always provide faulty grades. Garbage In, Garbage Out (GIGO).

So this nice, pat answer works for you science, math, and engineering types, where it easy to quantitatively evaluate what a student has mastered. What about the “softer” subjects? How do you evaluate a students ability to perform in an art class, or a music class, or give a speech? Once again, this is not a part of the discussion about using grades – it is a discussion about how to generate grades. And, once again, my answer is that someone teaching in those venues should be capable of evaluating their students using whatever methodology they find effective, and transform those results into a letter grade. A student who has truly mastered what the course was meant to teach would get a good grade. One who has not, won’t. What percentage of the course material has the student mastered?

Sure, a nice rambling narrative about the students style, strengths and weaknesses provides a ton of information, and gives the student a tool to improve their weak areas, but that is not the purpose of a grade. The details are irrelevant, and the student should have had the feedback and chance for improvement during the class – providing that information at the end of class (when the grades are finalized) isn’t doing the student any favours – either they already know what the teacher is trying to get across, or the teacher failed to communicate it during the class. The narrative bogs down the people that need the quick evaluation that a letter grade provides, and doesn’t really provide much additional usefulnes – if anything, it makes things harder – different people writing narratives about the same student will usually produce drastically different results.

I will not pretend that letter grades do not have problems. The variation between schools, courses, and curriculum all make it harder to know precisely what a given grade means. On the other hand, an accurate letter grade gives an indication of how much of the course material the student has mastered. This is a simple, effective method of communicating the students performance in a manner that provides an awful lot of information in a very concise manner. It does not provide complete details of the students performance, but it does provide a pretty good high level view of the students skills.

Note that this essay assumes that grades are not artificially inflated, curved, or otherwise doctored to make students, teachers, or school boards feel good. The premise is that a letter grade is direct indication of the percentage of the course material that the student learned. Not what the student learned compared to his/her classmates (curved grading), not boosted 20% to make the teacher or school disttrict look good (or to pass a certain percentage of the kids). If you are using grades accurately, you should occasionally get a class of all A’s. You should also occasionally get a class of all F’s. Usually, you should get a bell curve with the median solidly in the 73-76 range (that would be a  “C”). The kids folks might be upset that their kid earned an “average” when they know their kid deserves an “A”. but if you have well-established grading rubrics, and the appropriate records, your grades should stick.


4 Responses

  1. I went to a high school that had only two grades. Pass or fail. We rarely if ever had tests. We were not judged on exam scores but instead on portfolio review. There was no way to pass in my high school by “cramming.” You had to produce throughout the course of every class in order to pass. To me, this is the type model that the rest of our education system should move toward. Right now all you need to do to pass a class is memorize and regurgitate at test time. In a pass/fail portfolio review system, you need prove you can apply what you have learned.

    The scientifically impossible I do right away
    The spiritually miraculous takes a bit longer

  2. This has been a fascinating discussion for me.

    In my T/Th class, I have five A grades, seven Bs, four Cs, three Ds, and two Fs. In the Monday/hybrid class, four kids failed, four got a D, two got a C and five got a B – there were no A grades.

    I really do feel good about the way I grade my class – it’s pretty obvious from my distribution that I’m not doing anything overtly skewed – but I’m not content to stop investigating the topic. I think it’s important to continue to evaluate and question the way we do things, and I really appreciate discussions like this one to keep me in that thinking.

  3. reply to clapso:
    While you have indicated one ofthe bg problems with the current system, the problem that you are pointing out is a problem with test design. It is ppossible (and not that hard) to create tests that force the student to apply what they have learned. Cramming doen’t help in those tests – the student has to have mastered not only the information (cramming can help with that a bit), but also how to use the information. Only practice and application will get that internalized. Cramming won’t make any difference.

    While the portfolio and review process does provide a more detailed and individualized review of a students progress, you have not addressed the second purpose of a grade – to quickly and effeciently communicate the students level of mastery to another individual. That means some way of quanitfying the students mastery of the material in a way that allows a stranger to quickly compare the student to others. A portfolio and review or a pass/fail simply do not provide that function. If you really believe that the portfokio review and pass/fail system is the way to go, how would you suggest that a University screen and sort 25,000 applications in about 3 weeks?
    The portfolio review may be a better way to evaluate the student’s skill set to determine if the student is qualified to move on to the next grade – it certainly provides a much more individual evaluation ofthe students mastery of material. It dxoes not, however replace the communication of level of mastery that letter grades provide. Also, I would have seriouse issues about any high school that does not test at all (or hardly ever). The reality is tht if/when a student progresses to University, they will be facing tests. If the student has not mastered test taking, they are destined to fail no matter how much they learn….. It might feel nicer to not test, but unless tests are elimintaed from all levels of academia (includint SAT, GRE, etc.), any school that fails to teach its students to succesfully take tests is doing a disservice to those students.

  4. mrschili:
    sounds like you’ve got the grading nailed pretty well. your distributions look good, and the fact that you aren’t curving indicates that you really are measuring the students against a defined set of skills.

    How long have you been teaching this particular course? How long did it take to get the grading right?
    In my experience, the first few times I taught a coure, I would frequently have to adjust the grading up or down a tad, but once I had it nailed, things stayed consistant…

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: