A question on testing

In my online course, we were asked:

It has been said that the American educational system is obsessed with testing, and that students sometimes spend more time taking tests than learning. Teachers are often accused of “teaching to the test.” What is your view of the role of assessments in a course? Take into consideration the frequency of assessments, format, learning outcomes, levels of difficulty, areas to be assessed, and testing time.

I think that the concept of teaching to the test is mostly, though not entirely, a function of the accountability process put into place in our K-12 system. Most people who teach to the test in this environment are doing so with the expectation of teaching their students the fundamental knowledge which is necessary for them to pass on to their next stage of learning.

Sometimes this is essential and useful. What if there were no understanding of end outcomes for a course/class/year/school? Then each one would be different and a student from one would not have learned or even studied the same things that a student from another did. We would not have anything like a basic level that could be assumed within education.

And this can be useful for a teacher as well. I went to high school in New York, where every student in a course must take the statewide final exam in their course. It did several things for the teachers.

First, it removed the onus of “it was too hard for X” because it is a statewide requirement that you know how to prove that a line has 180 degrees (or whatever).

Second, the final was useful because it gave the teachers a clear set of objectives to be aiming for. World history wasn’t just supposed to talk about the history of a single country outside of the US, but was supposed to cover art and architecture and politics throughout world cultures. European history wasn’t just modern or early, but covered everything from the Etruscans forward.

Third, it allowed teachers a bit of flexibility in grading. This was not mandated (like it presently is in Dallas or Pittsburgh), but a student not doing well in a course, but TRYING, could be given a low passing grade with the clear understanding that the student would not pass the course if they could not pass the final. That was a statewide requirement. So a teacher could reward effort of a student without doing social promotion or effort promotion for the entire course.

But since we are all college teachers, the question should probably be applied to college teaching. I think that in college teaching to the test is problematic.

College teachers usually create their own tests. They know what is on them and they should certainly make the information available to the students through discussions, readings, lectures, activities, and assignments.

But students often expect a review handout that covers every single question on a test, rather than types of questions or types of information. If the teacher says that is what the review handout is for, then clearly it should do that. However I often see students not studying, but memorizing the review handout. Then, when they go to a test, they are upset because the question, as presented in the review, was not on the exam. That’s ridiculous. Why should you need to come to class if the review handout is all you need to know for the exam?

In addition I have upon rare occasions had a teacher who taught only what was on the test and nothing else. Each class period consisted of covering ten or fifteen minutes’ worth of material which was boiled down to a single question on the exam. And that was all the class consisted of. I am opposed to this approach.

Thankfully I am in English and for freshman composition, my most common course, the tests of all kinds are writing. Since the focus of the course is writing and the students are writing and the tests are writing, there is a clear confluence of testing and teaching coming together. (At least that is the goal.)

In my British Literature course, I primarily use out of class papers, where the students have to integrate what they learned in class and what they have found or discovered on their own. I do have two exams, though, which cover factual materials in a paragraph form. In those exams the students must write about the work addressing a certain question.

Because the course is involved and we do lots of reading, I allow the students to use their notes. They cannot use the book, but they can use anything that they created. This does two things. It encourages reading as we go along (because there is no way they can do all the reading before the exams) and it encourages them to take extensive notes on my lectures, our discussions, and the texts themselves.

Some teachers might feel that they ought to remember the issues on their own, but I feel that the way I have created the test allows me much more flexibility and gives the students a greater likelihood of doing well.

I don’t expect them to have memorized Beowulf or Paradise Lost. I don’t expect them to remember every description of the characters in Gulliver’s Travels. But I do expect them to have a general grasp of the work (both for the test and afterwards) and to be able to find the more specific information (either in their notes for the test, or on the internet throughout their life).

In addition, for the final, I ask the students to create questions based on the information we have covered in the course. I usually use at least two of those and if they come up with a question I was already planning on, I let them know that as well. …In a class with twenty questions, I will say, “Four of these will be on the test. Two of them were already on the test and two of them have been chosen from your suggestions.” I do not, though, tell them which those are.

One thing that one of my other colleges had and that I wish we had here was a final exam that was department-wide for freshman composition.

For the exam, there would be five or six writing prompts chosen by teachers and could not be prompts used in classes that semester. Every freshman composition student would take the exam. Then each teacher would grade two exams for every student they had in their classes, but they would not grade their own students’ exams.

These grades would be on a scale from 1 to 5. When the exam was first created and used, teachers were required to get two grades next to each other. So if one teacher gave it a 1 and another gave it a 5, someone else had to grade it and either give it a 2 or a 4 for it to pass. Obviously those with two 3s or whatever were looked upon most favorably. Because of the growth of the English department and the class sizes, however, that school now only requires that they be within two points. So a 1 and a 3 would be a match or a 2 and a 4. Then they would get the average as their final grade.

What this did was allow the students an outside, nonpartisan grader examining their work. It eliminated any discussion of bias on the part of the teacher (at least for the final) and it allowed the department to know about any outliers. If, for example, my student got a 3 (a 70%) on the final, but they had a 98 in my class, there might be an issue. Or, on the other hand, if I consistently gave all my students better or worse grades than they earned on the final, this was also clear when I turned in my grades. (We had to give the final exam grade and the final average.)

This also let us know of any grading outliers in the department. If someone consistently gave significantly better or worse grades, their averages could be looked at and they could have extra norming help.

It created an interesting experience once when the first two graders gave a paper a 1 and a 5. The 1 was given because the audience (the teacher) was told that English teachers were idiots and the whole paper focused on how much smarter the student was than English teachers. The 5 was given because there was not a single grammar error in the whole eight pages of the paper (written in two hours). When other readers also failed to come to agreement, the entire department got back together to grade and discuss the paper.