This discussion of why letter grades vs. a simpler point system reminds me of how one is evaluated as he/she progresses along the education system (at least in the U.S.). First you get letter grades (or some form of an ordinal level of measurement) in elementary to secondary school. And often these are converted to equivalent grade points. However, certain numerical scores become more important – like the PSAT, SAT, GRE, GMAT, MCAT, LSAT – as one climbs the academic ladder. And once in grad or other professional school, letter grades are pretty much discarded. Sure, they may still be assigned, but they don’t mean much and it’s rare that anyone drops out because of grades. A simple pass/no pass system is often employed. The graduate level student is more concerned about a passing numerical score in the qualifying exams, the medical Step exams, or bar exams. At this stage of education, letter grades are essentially marginalized because the results need to be as objectively determined as possible. When the results will decide if someone will be minted a Ph.D. or will be allowed to practice medicine or law, we want hard numbers and not soft letter grades. Based on all this, my conclusion is that letter grades are used mainly to allow more subjective academic assessments and are retained for educator convenience and easier interpretation by the reader.
There are no numbers involved in deciding whether one earns a PhD. How could you numerically grade something as individual as a thesis?
minor quibble
isn’t a A- 89.6% to 89.9%? most teachers I had round up. they’d call it an A- or sometimes just an A
The answer, of course, is that the letter grades came first, and are expected by the general populace. A report card that reported class score averages would leave the parents confused. It would also fail to convey much useful information, since absent some conversion of the numerical score average to the letter grade, each teacher’s average score might mean completely different things.
Of course, in reality, they do anyway (getting an “A” from me is not as easy as it is to get one from other teachers in my high school, for example). But at least a parent of a student who has received an average score of 95 from me knows their son/daughter falls within that class of students I adjudge to have done "excellent"ly well. There is, of course, a certain amount of lovely, circular logic to all this.
Just how silly all this can be is reflected by states which change the conversion scale for percentages to letter grades. My state, for example, uses a conversion scale of 93% to 100% = “A”, 85% to 92% = “B”, etc. The theory behind this increased floor for each letter equivalent is that it will force students to do better work to get the same letter grade. But there is no guarantee of this at all. There is no guarantee that the numerical score represents a fixed percentage of correct answers on a standardized set of assessments. So a teacher is free to assign any numerical grade they want to any assessment result they want, making it impossible to establish an equivalency between a 91 under the old system and a 91 under the new system. In practice, most of my colleagues assign essentially the same percentage of “A”, “B”, etc. grades as they did before the state made the switch.
The better question, of course, is “Why do schools assess semester/year performance with a simple letter grade?” Anyone who has been involved in high school from either student or teacher standpoint knows that one person’s “A” is not neccessarily equivalent to another’s “A”. Some get an “A” because their assessments show a very high understanding of the material taught; others get an “A” because they are dilligent in accomplishing the work required by the teacher in a timely fashion. Since the meaning of the letter varies by so much from student to student, and teacher to teacher, it would seem that some more intricate assessment method would serve the purpose of establishing how a student has done over time in a course better. But, of course, then you wouldn’t have a GPA that could easily be compared to other students to establish a pecking order. </sarcasm>
Letter grades and numerical grades are both just scales of symbols upon which one can hang a student’s performance at various points; it isn’t really about letters vs. numbers (although the choice to use numbers does further carry an unearned connotation of objective, cardinal measurement…). It’s about how coarse or fine you want the evaluative scale to be. And often, very fine-grained scales give a misleading impression of phantom precision, while coarse-grained scales better capture the degree of meaningful evaluation.
You might well ask why we give numerical grades out of 100 instead of out of 10000, with the extra two digits of precision. For almost all purposes, worrying about the extra two digits is just adding noise on top of what was already pretty noisy, that’s why.
Also, on preview: I feel DSYoungEsq’s post is very good and deserves acknowledgement as such.
i too didn’t why is it so?
The problem is that there ISN’T a standardized grading system…my 2nd grader kids get report cards with V, +, n/a, C, dolphin.
(Joking on the dolphin, but honestly, the rest is on their reportcard)
At the same time, they can show me a chart where the kids are in reading comprehension on a words per minute basis, and where they ‘should’ be, compared to a statistical average. They can show what phonemes they’re deficient on, and teach the kids on a ‘if you don’t get it this time, you’ll be seeing it again in 8 months’ plan.
MAkes it hard to tell if the kids are alright on one hand, but on the other, they seem to be seriously over-diagnosed…what I’m not sure of is how it all falls down on the state standardized assessments. Does B equal ‘orange’ equal ‘passes and will be a good unique flower in society’?
It isn’t tradition, its logic (though I don’t agree with it). As I have been told by education professionals, letter grades better reflect what a teacher believes the student actually accomplishes. Numerical grades give a false feeling of precision. Students (mostly parents) will feel that there is a difference between a 86 and and 88. In reality the accuracy of the measurement is probably about 10 points. At any rate, the accuracy is nowhere near 1 point. Reporting more digits than are significant is bad reporting. And the kids don’t benefit from false comparisons.
It is a pretty compelling argument, but I have always been in favor of numeric grades. Perhaps they should be reporting a standard deviation with each grade as well.
In my experience (graduated HS in 2004), we used both - the grades were presented to us each 9 weeks as a percentage, but then on our transcript, all we had were A/B/C/D/F for the semester. But since we knew the formula for computing semester grades (.41st 9 weeks+.42nd 9 weeks+.2*final exam), it let me know how much to care (or not care) about finals.
Well, anecdotally, I know for a fact that a least one teacher draws little doodles in the grade book. Then, at the end of the semester, when he sees the “Happy Llama” or the “Hipster Smiley” he knows that that kid did an exceptional job in class participation that day.
It’s even gotten to the point where kids’ll ask “So, if I add something totally new to this discussion, is that worth a Dancing Possum in the grade book?” Or someone’ll run in late and ask “Shoot, does this mean I get a Smelly Roadkill in the book?”
The sad thing is that this is college.
(One kid brought me a Breakfast Burrito… wrapped in an illustration of him handing me a Breakfast Burrito and me handing him a Smiling Raccoon…)
The problem with precise grades - 87.6% or 53.9% - is that they ahve to be based on precise marks. Sum up “5 tests, 10 homework assignments 2 midterms and a final… divide by 270”.
I recall kids from my classes who would whine to the teacher over every assignment, trying to get an extra mark out of 20. The more steps in the scale, the easier the whining becomes. As has been mentioned above, how do you tell 88.9 from 91.3? OTOH, A vs. B is an easy call. Not every test can be as structured and precise as the SATs.
The problem with letters is the moving target. I always thought it was F: <50, D 50-60, C 60-70, B 70-80, A80-90, A+ 90 and above. If every location has a different scoring system, who cares as long as A means excellent and F means flunk.
And - what do you do with a creative and well-written essay where the student consistently mispelled several words? Are you marking creative communication or grammar and spelling? Or a happy medium? Is your marking scheme fixed or impressionist? Every teacher has their own style, and in the long run, the question is: can the student read, write, and figure well or poorly?
Wow, and I thought I was odd for drawing smiley faces on perfect assignments… It comes of being the son of an elementary-school teacher. If I had glittery star stickers, I’d hand those out, too.
Well, it’s certainly not a letter grade that you see on your thesis! I would argue that it’s strictly pass/no pass at that point in the program, with your dissertation advisor being the primary arbiter. And numbers do decide if you pass qualifying exams at many schools (note: I did not say all schools) and those are the first hurdles to surmount as you climb the Ph.D. hill. That’s when the weeding first starts. No need to even think about writing a thesis if you can’t get by the quals or prelims.
One advantage of letter grading is that it can be normed to the cohort, rather than in absolute terms.
What you are doing is comparing students against their peers because you can adjust the standard required for each letter, and some of the adjustments can also allow for differences in exam papers, since these are not always consistant.
Letter grading is really more comparative grading and can be independant of exam scores, this can be useful in subjective studies, such as humanities where one is measuring the capacity for debate and criticality as much as for the background research.
Another anecdote to support the subjectivity of grading, esp. in the “soft” sciences:
A student asked me why he got a B on an assignment. I said “remember the examples I showed in class? Those excellent ones were A’s, and the average ones we looked at were C’s. Where would you rank yours?”
“Well, I guess, in between those…”