Saturday, November 26, 2011

Grading Systems

I have a beef with the common grading systems in US education. My smallest issue is that we are continuously allowing it to slip. Originally, the grade of a C represented a satisfactory understanding of a subject, while grades of B or A represented an exceptional understanding of a subject. In oriental countries, this is still true, but not in the US. If you perform only to the expectations of a class in the US, you get an A. To get a C, you must perform well below expectations. Recent "reforms" in education have made the situation worse. Teachers in public schools now must advance poorly performing students, even if they barely understand the subject at all. This is destroying the already poor reputation of the US public education system. My primary problem though makes this one pale in comparison.

Would I be wrong to say that the intent of grading is to represent the proficiency a student has in a particular subject? When you apply for a major in a college, especially when that major is for a post graduate degree, they look at your previous transcript. If the major is especially competitive, they look at your grades in specific classes related to that major. A student attempting to enter a post graduate math program might have a strict review of grades in previous college math classes. The reviewers make the assumption that the grades listed on the transcript represent the student's proficiency in those areas of math and in math in general. This is the same for any subject. There is, however, a problem with how our education system gives grades and how it interprets them.

While it is assumed, when looking at a grade, that it represents proficiency, it is not actually true. Teachers and professors may take any number of things into account when giving grades. Class grades are generally composite grades from all of the assignments given throughout the course of a class. This means that a student that takes an extra two weeks to understand something which is taught, but then catches up, will have a poorer class grade than a student who has identical proficiency at the end of the semester, but who did not have difficulty early on. A student who has some special circumstances which cause homework to be late, but who understands the material quite well will end up with a class grade that under-represents proficiency. A professor might burden a student with excessive, unnecessary busy work that the student does not have time for, which would also cause the class grade to misrepresent proficiency. In addition, some college professors like to give assignments which are not related to the subject at all, but which may help some students learn better. Again, missing these assignments may not affect a student's proficiency, but the student will still be given a grade that indicates a lower proficiency than actually exists.

I have dealt with many of these situations. I once took a physics class that required weekly workbook assignments, in addition to regular homework. The workbook assignments were generally ten pages or more and often took two hours a day, five days a week to complete. The weekly homework required only two to three hours a week. The workbook assignments were excessive and unnecessary, especially given that this was my second time taking physics (the first class did not transfer, but I had gotten an A). I was also taking several other quite difficult classes, so I chose to forgo the workbook assignments so that I would have time to study the other subjects, which I was not so proficient in. Ultimately, I got a class grade of C in the physics class. When you look at my assignment grades for the class though, it is obvious that a C did not accurately represent my proficiency in the subject. I had many scores of zero, where I had not turned in assignments, however, regularly interspersed between the zeros were grades between 93 and 100. There were no grades lower than 90 for anything I had completed. As I had missed only a very few homework assignments, and the tests and homework assignments were fairly comprehensive, my proficiency in the subject was quite well represented by the grades of the assignments that I had completed, but my class grade indicated that my proficiency was similar to that of a person that had learned only a little over half of the material.

My first two physics classes were very different. My professor gave final grades either as composites of all assignments including tests, or only the grade of the final, which was comprehensive. He would choose whichever grade was higher. The reason he chose to grade this way is that it more accurately represents the proficiency of a student than just the composite grade. When he explained this, he also pointed out that the occurrence of a final grade higher than a composite grade was very rare. My second semester of physics, my composite grade before the final was a B. During the last two tests, I had kept careful track of the things that I did not know well, and spent extra time studying them. I scored a full 100% on the final, showing high proficiency. Instead of getting a B, or B+, which would have indicated only moderate proficiency, my class grade was an A, which accurately represented my proficiency in the subject.

Now, I am not saying that all classes should be graded like my first physics professor chose to grade. That system also has its flaws, though not nearly as bad. First, I think that teachers and professors need to carefully consider how to grade so that the grades accurately represent proficiency, with as few taints as possible. I realize that it is impossible to grade entirely accurately, but some effort should be expended to minimize inaccuracies. If a school feels that it is important to also keep track of a student's dedication, work ethic, and punctuality, then maybe a second grade should be given and tracked, but this should not be allowed to taint the representation of proficiency of the class grades.

I propose, however, a system that makes such changes entirely obsolete. A less formal milestone based system would far more accurately represent proficiency than a class based system. Instead of collecting letter grades and then using them to compose a GPA, progress could be tracked more directly. Each subject would be divided into a collection of significant milestones. The subjects of a single class would be divided into at least 5 milestones, though some classes might be divided into 10 or more milestones. Instead of using a GPA for gauging progress, progress would be determined as a function of milestones over time. A full time student might be expected to complete at least 25 milestones in the time of a normal semester, or 50 milestones a year. There might be a requirement that a certain number of those milestones be related to the student's major, or other required classes. A student could exceed that number by any amount. Using milestones, letter grades would be meaningless. Each milestone test would be pass or fail. A failed milestone test would not show up on a transcript, because the lack of a passed milestone test already indicates a lack of proficiency. Accurate grading is not the only benefit of a milestone based system.

A milestone based system allows students to work at their own pace. A milestone system does not even need regular semesters. A milestone system would be ideal for those with jobs, who do not have time to attend school full time. A milestone system would also dramatically benefit people like myself, who are already quite proficient in the subject of their major. I could easily finish all of the milestones associated with the first two years of my major, in only a few months. This would reduce the cost of my schooling by nearly half. I am sure there are plenty of others in a similar situation. A milestone system would also help those who have difficulty with certain subjects, since they would have longer to study and would be able to spend more time on the parts they find difficult and less on the parts they find easier.

A milestone system would also be easier on teachers and professors. Instead of regular classroom sized lectures on a series of subjects in succession, ignoring the needs of the students, professors could give lectures to larger groups, spending more time on subjects that need attention. Instead of a class schedule, professors would have lecture schedules. Any student could sign up for a lecture and attend, and if there was room, additional students could attend without signing up. If no students signed up for a specific lecture, the professor could instead do a question and answer session. Since professors would be doing larger lectures less often, they would have more time for one on one work with students who need it. Also, in a milestone based system, students would be more responsible for their own learning. This means that there would be almost no graded homework. Students could ask a professor for feedback or help on a specific assignment, but automated systems could be used for basic feedback. The only real graded work would be milestone tests, which could be graded by paid graders, leaving professors more time for helping students who need it. This would allow professors to have much more time for teaching by reducing the task of grading and placing what is left on less skilled workers, who do not need the expertise of the professors to grade.

I believe this sort of system would dramatically improve the performance of our educational systems. It would reduce the workload of teachers, giving them more time to do their primary job, teaching. It would remove the inherent discrimination against slow learners and the learning impaired that exists in an arbitrarily timed class based system. It would also remove the discrimination against fast learners and those who are already self educated, by allowing them to complete milestones at as fast a pace as they desire. In addition, it would overhaul our current grading system to accurately represent proficiency in each subject, because each milestone would be verified for each student. I also believe that it would dramatically improve education, because no student would graduate without having passed each individual milestone required for their major. No student could get by just barely.

The primary argument I have heard against this sort of system is professors who say that because they had to work through the current system, everyone else should have to. This is a lie. By this argument, I can show that those very professors should be doing hard labor in the fields of a farm for the own survival (or even hunting and gathering naturally occurring foods), without any modern tools or other conveniences. Those who claim that the next generation should have to work just as hard as they did do not deserve the conveniences of modern technology and medicine. The progression of the human race requires advances in education. Even very specialized fields require more knowledge than the average human can learn in only 4 or even 8 years of schooling. As technology and science continue to advance, this will only increase. If we do not find ways of increasing the speed and efficiency of learning, we will find that technology and science advance ever more slowly. Eventually, learning enough to continue to advance will take a lifetime, and at this threshold, humans will no longer be able to continue advancing, because people will die before they have learned enough to advance further. Computers have already expanded that threshold, but the advancement of computers itself is subject to that threshold. If we do not improve education, the human race will eventually reach a technology cap, and cease to progress any further. Improving education, even a little, may allow us to create technology that further extends that threshold, and may even allow us to extend it indefinitely. If we choose not to improve education, because "it was good enough for us, so it is good enough for them," then we are taking an enormous risk. Let us not lie to ourselves. The current US eduction system is terrible. It does not fulfill its purposes. A milestone based system would allow students to prove their merit by their rate of learning, instead of being forced to conform to a rate of learning that is optimized only for the most average student, and that makes fast learners hate school for the boredom and slow learners hate school because they can never learn fast enough to understand anything.

Eventually some country is going to adopt this system. That country will very quickly begin to excel in technology and science at rates never seen in human history. We are already behind nearly every other 1st world nation. Shall we allow Japan, or China to beat us to the punch, and become the new greatest power in the world, or will we do it first, and reclaim our dominance in technology and science?