Our broken testing system

I've often participated in group grading.  My first experience with it was in 2007, the year I graduated from BYU.  It's horrible.  This semester, as I was sitting with a couple other people grading a common final, I reflected on the futility of the final exam system that we have in our colleges.  (Of course, the standardized tests we have in elementary schools are far more ridiculous--trying to determine a teacher's effectiveness by asking children to draw bubbles on paper, but I won't go there today.)

What is the purpose of a test?  Why do teachers test students?  If you ask me, the purpose is to determine how well the student has internalized the material taught in the course.  The purpose should be to illuminate the level of understanding (or lack thereof), to decide whether the course has been beneficial to the student.  When I write a test, I want to make sure that I test all of the key points that I taught.  I want to make sure the questions are worded unambiguously, since I feel that trying to understand my psyche would make the test far more difficult than it already is to most math students.  

But the real problem isn't in how the tests are written.  For the most part, I think we (teachers) do a good job of writing tests, and of indicating which points are to be tested most heavily.  The problem lies in what we do after the student takes the test.  In my opinion, the current method is wholly ineffective for any positive end.  What is typically done is the tests are evaluated, points are assigned, and a grade for the course is given.  Why is this ineffective?  Because the student never sees the test.  

When I give midterm tests, I always make a big point to give feedback as appropriate on the test and to go over key problems in class after the test is returned.  I think that a test should not only be an evaluation of how well the student learned, but also a key part of the learning process itself.  That is, students should learn from tests.  When I grade a student's test, I am not making an effort to see the maximum number of points I can subtract from the total, nor the maximum number of points I can in good conscience assign.  What I am doing is indicating to the student the difference between truth and error on eir test.  If a student writes something that is false, I feel that it is my duty to so indicate.  And I feel that I am doing the student a favor by doing this, because then the student is able to take the feedback and learn from it--to find the things ey has to review, to weed out the false ideas that have implanted themselves in eir head, and to more accurately approximate the truth being taught in the course.

The final exam structure in college does not allow any of this to happen.  The student never sees the graded test.  It is impossible for em to learn what mistakes ey made.  The teaching moment is destroyed.  

One other major problem I have with the grading system is that we teach children--from the very young age of 5 years--that being wrong is bad.  A paper that is less than 100% correct is undesirable.  Students who are wrong are stupid.  Smart students are the ones who never get things wrong.  This is false.  And it needs to stop.  It is the reason why students are afraid to answer questions in class--the fear of being wrong is too great.  We need to teach students that it is okay to be wrong.  In fact, it is good to be wrong.  We need to encourage students to be wrong because being wrong is how we learn.  No one knows everything.  No one does everything without mistake.  And we shouldn't expect that of people.  We should expect and encourage people--especially in the classroom--to make mistakes.  

This is the main reason I do not like grading papers.  To me, a student's grade should not be based on eir performance on tests, but rather on the amount that a students learns.  In fact, if I had my way, there wouldn't be any grades at all.  Every class would be pass/fail.  I would give my judgement on whether I felt the student's understand of the material was adequate for em to move on to the next level, and that's it.  

To me, the ideal would be to give the test back to the student with all of the markings indicating what is right and what is wrong, but with no score attached.  I know what it's like to be a student and to feel like my worth is determined by a number written in red ink at the top of a paper.  We shouldn't do that.  We should encourage students who make many mistakes on the test, and allow them room to improve during the semester.  Most importantly, we should make the final exam before the last day of class so that we can return the graded tests to the students and help them to understand the things that they were struggling with.  

Of course, ideally, teaching should happen one-on-one.  With 30 students, I can't know how everyone's doing at every point in time.  I can't know which topics teach in more detail and which ones I can gloss over more quickly.  I can't tailor the material to the student.  With one student, I can do that.  I can get immediate feedback from the student on how well ey is understanding what I'm saying.  The student doesn't feel social pressure to remain silent, and is much more likely to voice uncertainty about a topic or ask a question.  It would be nice to be able to sit down with every student individually and discuss eir test with em.  And I do have the comfort of knowing that this is how the learning process happens at the graduate level.  I have an adviser and he can give me feedback on whether the work I'm doing is good or bad.  It would be nice to see more of that kind of learning--mentoring, rather than lecturing--happen in lower levels as well.