I received an email question recently from a professor at Rutgers about testing. What students think about their teacher is heavily influenced by how well the testing and grading process is managed. Here is an edited version of my response to him.
In my new (free) teaching book (Transformative Education), I talk about testing and grading as the last of the four essential points in my "Education Diamond." It cannot be an afterthought but rather a key part of building a successful learning experience for your students. Anything that students think about as much as they focus on testing and grading needs to be taken very seriously. All tests should be motivational. They should be fair. They should serve as a warning. They should be challenging. They should not be impossible.
I do not use test banks. Anything as important as a test should not be turned over to some stranger who does not know your goals or your students. I know learning to write good questions is hard but with a bit of practice you can become an expert. Have some courage. Write your own questions.
When my students walk out of the classroom following a test, I want each one to think, "Those were reasonable questions. If I had done everything Professor Hoyle had suggested, I would have done well. The test was not easy, but it certainly was not impossible."
When I create the tests, I have several rules that I explain to my students. Clear communications are an essential part of my class.
--Anything we cover in class is subject to being tested. Conversely, if we did not cover a topic in class, it will not be tested. If it is not important enough to mention in class, it is not important enough to test.
--In writing test questions, I try to make 1/3 of them very challenging. I call these “A-level” questions and they focus on critical thinking. I want to measure the true level of understanding that a student has attained. Another 1/3 of the questions are “B-level,” they require understanding, but the level is not as in-depth. The final 1/3 (the “C-level questions”) are easier—more inclined to require memorization and less in need of a deep understanding. After five decades, I have found that having three levels of questions is the best way for me to assess each student’s actual understanding of the material.
--Each question must be tied to something we did in class. The tests and the class must be connected. As students read the questions, I urge them to identify that connection and consider what we did in class. I also suggest that they write down potential questions immediately following every class to consider how topics might be tested. When the test comes around, that list is extremely helpful in spotting the topics we have stressed.
--Each test question requires careful reading. Students are often nervous and will tend to skim questions and then be shocked when they make silly mistakes. I urge them to take a deep breath and read the questions carefully and completely before jumping to an answer. Underlining or highlighting are helpful.
--In writing the questions, I do not have the best student in mind nor the worst student in mind. I usually aim my questions at the student who is about 1/3 of the way down the list of students, a student who typically makes a B or a B+. If your overall test aims too high, you really do leave a lot of good students feeling stupid. If your overall test aims too low, many students will make an A without deserving it.
--Almost without exception I do curve every test. It seems to me that it is my job to decide the difference in “excellent” work (an A) and “good” work (a B) and “average” work (a C). I correct every test and then line them up from top to bottom. I identify what I believe is the lowest “excellent” test and what is the lowest “good” test and what is the lowest “average” test an then I create a curve to establish that as their grade. Some of the curves are rather convoluted but that happens less often than you might think.
--In my classes, 15 to 30 percent of the students make an A. I don’t intend for that to happen, but it seems to result on a consistent basis. “Excellence” is not easy to achieve in life. I realize grade inflation is prevalent, but if you explain that you are looking for excellence, most students know in their hearts whether their work has truly been excellent. I have said this before but giving an A for work that is less than excellent actually hurts the student and sets a bad tone for the class.
--I provide an answer sheet that I email the students immediately after the test. At that moment, most students are keenly interested in how they were supposed to answer questions. I want to take advantage of that interest. Also, by typing up the answer sheets in advance, I find typing errors and questions that are easier or harder than I had expected. I can fix a lot of testing issues in advance by typing up the answer sheets.
--Occasionally, I write tests where I give the students all the questions in advance but with specific variables omitted. For example, I might write, “Use XX interest rate for this problem and assume a life of XX years.” What I have found in such cases is that the test becomes a real learning experience. Students work twice as hard preparing for a test when they have a valid idea about the nature of the questions.
--I tell students that I will always be glad to talk about their tests and how to improve. However, if they want to argue a test grade for any reason, they have seven days. I don’t want a student coming in two months after a test wanting to dispute a test question that I can barely remember. I don’t mind a student arguing a test grade. It is their grade forever. They have a right. However, they must submit their dispute in a typed form. I usually take several weeks to respond because I want to take my time. Whether I add points or not, I type up my response to explain what I did and why. By formalizing the process, I have found that students only dispute a grade when they truly believe they have a valid argument.
--To try to keep students from getting too stressed out, I refer to my first test as “Spring Training.” It does count but it is primarily a chance to see how well they are progressing and what adjustments they might need to make. I want learning. I am not trying to create undue stress.
--About 24 hours after my first test of the semester, I send out an email that I title, “When you get your first test back, if you are not happy with your grade, here are some suggestions on what to do.” I then give them as many suggestions as I can. At that point, many of them need some guidance. For many years, I have argued that the few days after the first test is the most critical moment of the semester. Almost immediately, students tend to start getting better or getting worse. As a teacher, take control of those days and push the students in the direction that you want.
I could probably talk forever about testing and grading. But, I’ll leave you with one last comment. I want my students to learn to develop their critical thinking skills. I tell my students that all the time. However, if I don’t test the development of those critical thinking skills, the students will simply assume that what I really want is for them to memorize the material. Keep that in mind as you write those tests.
As always, if you want a free download of my new FREE teaching book (Transformative Education), it is available at the following URL. In addition, if it is helpful, please share that URL. Educational ideas should be shared and not hoarded. (Thanks to so many people sharing the URL, the book has been downloaded on the average every 3 hours and 50 minutes every day for the past 136 days. That is a lot of people working to become better college teachers. THANKS TO ALL.)
https://scholarship.richmond.edu/bookshelf/375/
ADDED
LATER: POP QUIZZES
After
reading the above essay about testing, a professor at Troy University asked for
my opinion about pop quizzes. As
everyone probably knows, I am not shy about giving my opinion about anything
and everything in college teaching.
I believe it was the spring semester of 1982 and I decided to start giving pop quizzes to one particular class because I really wanted the students to come to class well prepared. I am a strong believer in experimentation so that was simply an experiment. I never did it again. That probably shows what I thought of my results. Why have I not returned to pop quizzes?
--For the entire semester, the class had a tension to it that I did not like. The students arrived every day wondering whether “today is the day.” Tension is not necessarily bad for a class, but that type of tension seemed counterproductive.
--On days when I did give a quiz, the class always did poorer after the quiz. The five-minute quiz got their maximum attention and, when it was over, they exhibited little interest in the remaining conversation. It was like they had run a 100-yard dash and needed to rest.
--The students became too interested in gambling. They would assess the odds each day of a quiz and adjust their study accordingly. I wanted them to be well prepared every day and not just on the days when they suspected a quiz.
--However, my main reason for abandoning pop quizzes was that the only reason to do them was to push students to prepare for class. I believed that I could get them to be prepared without holding the sword of a quiz over their heads. In my book (Transformative Education), I write pages and pages about using puzzles (and, for me, cold calling) to encourage and intrigue students enough to be prepared. I believe the biggest challenge in college teaching is getting students to prepare properly so I work very hard to create questions/puzzles that will intrigue them into being ready to be part of the conversation. In fact, I sent my spring students their first puzzle three weeks before our first class just to pique their interest.
I am sure that some of my readers will have found ways to make pop quizzes a positive part of their class experience. However, for me, it was a total negative and I have chosen to address the problem of preparation in other ways.
