Joined at the Hip: Teaching and Testing
I enjoy teaching but dislike giving examinations and grading. Is there something wrong with this picture? I became a college teacher to educate students, not to spend time deciding whether an essay answer is worth a B or a B- or whether average exam grades of 87.3 are B+ or A-. For me personally, grades are a secondary and derivative issue at best, an anguished responsibility at worst. My dislike of grades has led me to search for ways to separate them from my teaching. I don’t say “everybody's teaching" -- only from mine. If other instructors enjoy testing and grading, I won’t try to take away their fun.
Under some circumstances, my teaching and testing have been separate. When I give a guest lecture or workshop at another college I arrive in the morning, meet with classes in early afternoon, lecture at four, and depart without giving an examination or grade. Is this teaching? It’s fuzzy, but most people wouldn’t think of it that way. A stronger case for inclusion under the teaching rubric can be made for short courses and workshops. None of those I have given or taken has included tests or grades. There are also non-credit adult education courses without tests, grades, or certificates. People enroll for enjoyment, intellectual stimulation, and skill enhancement. Tutors and coaches instruct individual students without formal evaluation of performance.
My objection is not to all testing, only to summative (end-of-course) testing for an official record. I know the disappointment, hurt, and dashed hopes that a low grade brings. However, I fear that awarding high grades for poor performance will produce unrealistic expectations likely to be dashed later at even greater cost, and also reduce the value of a diploma. Society pays me to sort students by the quality of their work.
I have dreamed of a different system, in which I am hired specifically to teach without being required to grade. I would be paid for giving a fixed number of lectures and my responsibilities would end at that point. Grading would be the responsibility of an agency other than the instructor. This is the case with the Advanced Placement examinations. Under this system, students and instructor work toward a common goal, to insure that all students perform well on the exam, as this reflects well on both instructor and students. This is indeed “teaching to the test,” which seems a reasonable approach when the test is a valid measure of student performance. State or national exams like this, developed to yield separate scores in different core subjects, could sever many existing linkages between teaching and both testing and grading. In core subjects, teachers can focus on covering what is on the exit exams, and assign grades based on scores on the exit exam administered by the state agency. For elective subjects, teachers would continue to test and grade as they do now.
I leave to another forum the question of whether any single examination can properly assess critical thinking, creativity, motivation, and other valued intangibles. My guess is that this is difficult under the best of circumstances and more so under the stress and other constraints of a high stakes examination. I doubt that my exams in large classes do this. Most likely they test some combination of general intelligence, prior academic background, note-taking skills, test-savvy, and rote memory, which are all qualities valued by society, but not what motivated me to become a teacher or what I most value in students.
A system similar to the AP testing approach would be used in large introductory college courses. The examinations could be constructed by an intercampus consortium, a professional society, or by a testing service. Instructors would be paid extra for monitoring and scoring examinations, not for their own classes, but exams taken by students taught by other instructors. This arrangement separates teaching from summative evaluation. Instructors can still use diagnostic tests and projects as needed throughout the term, but are not responsible for assigning grades on the summative examination.
I began college teaching under a system like this. It was the most miserable teaching experience of my life. Along with other junior faculty, I was assigned a section of my department's introductory course, in which all instructors used the same textbook and gave the same final exam scored by machine. The final examination consisted of 100 multiple-choice questions selected by a committee from the test bank supplied by the textbook publisher.
When I found out about this system during my first teaching year (as is typical, no one mentioned exams before I accepted the job and I didn’t ask), I questioned my role in the classroom, and felt obliged to share my reservations with the students. With a final exam based strictly on the textbook, I doubted the value of students coming to class, other than to seek clarification when something was unclear, and this would require a question-and-answer rather than a lecture format. I suggested to the class that it might be more valuable for them to stay home and read the textbook. Attendance at my lectures dropped by two-thirds and, at the end of the course, I received the lowest student evaluations of my career. However, there was no indication that the performance of my students on the final examination was any worse than that of other instructors.
I was demoralized by this experience and vowed I would not teach again under this system, in which instruction appeared to be separate from evaluation. I say “appeared to be” because, after the course ended, another instructor informed me how he and his colleagues handled the situation. In addition to the standard final exam, there were midterm examinations that contributed to student grades and these were composed and graded by individual instructors. By including material in the midterms not covered in the textbook, students were motivated to attend class.
In retrospect, I don’t believe I disadvantaged my students by suggesting that they stay home and read the textbook. I still came to class even if most students did not, and was available to answer questions and elaborate on points in the textbook. The basic factual information, including major theories, concepts, and research findings can be (or should be) provided in a textbook. This does not eliminate the need for instructors but changes their role. Instead of duplicating material in the textbook, they can lead discussion sections and supervise laboratory, studio, or workshop sessions, besides being available to answer student questions. What is eliminated in this system is lecturing on material already in the textbook or that can otherwise by provided to students online or in hard copy.
I put aside this issue for several decades, and graded students in my large classes using multiple choice tests scored by machine. Unlike my miserable first teaching experience, I wrote most of the questions myself, which covered both textbook and lectures. In small classes, I evaluated students on the basis of term papers, projects, participation, and essay examinations.
My grades were remarkably similar to those given by other instructors teaching similar courses. I knew a few faculty members who challenged the status quo by giving blanket A’s or allowed students to self-grade their work. Most of these mavericks did not last long in academe. In any event, I could not see such a system working in my large lecture courses. There would have been sanctions if all 250 students in my class awarded themselves A’s. Indeed, as a department chair, I would have recommended sanctions if an instructor gave blanket A’s in a large lecture course, in the belief that this would undermine the integrity of the grading system and thereby demean the value of a diploma.
In small classes one can replace grades with written narratives, if anyone cares to read them, but this will not work in large classes. The pros and cons of eliminating large university classes go beyond grading issues and deserve consideration by themselves. Another solution is to abolish grades entirely and make attendance the sole criterion of course completion. Graduation becomes a matter of having attended a specified number of lectures. I do not believe that a diploma awarded in this system would have much credibility.
I am glad I have not forgotten the stress of examinations and grading. I did not like this stress when I started teaching and I do not like it now. I will not abandon a search for alternatives, but am resigned to teaching under a system in which instruction is paired with testing and grades.
Bob Sommer is distinguished professor of psychology emeritus at the University of California ad Davis.
- A Defense of Textbooks
- A Curricular Innovation, Examined (Part 3)
- CourseSmart announces analytics program to measure student engagement
- Why Digital Supplements Drive High-Priced Textbook Adoption
- Essays on the flaws of peer grading in MOOCs
- What History Students Read
- Professors Gone Paperless
- A Curricular Innovation, Examined
Search for Jobs