Starting with my second year of full-service teaching, namely, the academic year of 2010-2011, I began a practice that I continued for the next nine quarters that I taught. This was a practice of MCQ-based class quizzes. On this page, You can access my MCQ class quizzes by looking at the information for individual courses I taught linked to from my teaching page. The most recent, and perhaps most sophisticated, round of quizzes were prepared for my Math 196 (linear algebra) course. If you want to see one example, check out this quiz.

I’ve written more about general aspects of multiple-choice questions here and about class quizzes here. On this page, I focus on my personal experience using MCQ class quizzes and why I think more instructors should consider using these as a substitute or complement for other methods of diagnosis and evaluation.

The inspiration for MCQ class quizzes was an ACT preparatory class I taught for high school students in the Summer of 2010. I made a mess of the ACT Prep class. However, I did experiment with MCQs somewhat during that class and this led me to the belief that I should experiment with them for my regular teaching work as well. I was also able to learn from some implementation mistakes I made with MCQs with the ACT prep class.

Design elements

The following are some aspects of the quiz design.

• Format of administration: The quizzes were administered in sheets I would print out and give to the students. Each sheet contained a space for the student’s name at the top. Each question began with the question statement, followed by the answer choices, followed by a blank space where the student could enter his/her preferred answer choice.
• Availability of quizzes: Each quiz was given to the students at least one class prior to the due date, regardless of whether it was an in-class or a take-home quiz. In addition, the quizzes were made available online through the course management system used for the class. I also provided spare copies of quizzes to students who missed the previous class and needed to submit the quiz in a given class.
• Submission flexibility: Students were generally expected to turn in the quizzes in the sheets I gave them in class. However, students were free to email me their solutions or hand their quiz attempt before the time of class if they anticipated missing class. This means that students who missed class because of sickness or scheduling issues were still expected and able to turn in solutions (though the low weight assigned to class quizzes overall in the grade — 2% of the total — means the student doesn’t need to worry too much). Students were also allowed to spend a few minutes (up to 0-30 minutes) after class completing the quiz, or to email me their solutions an hour or two later if they had exceptional circumstances.
• Solutions: I would put solutions online (on the course management system) a few hours after the quiz was due and include a review of student performance. I returned the graded quizzes to students in the next class or whenever I next met them.
• Take-home versus in-class: Some quizzes were take-home — students were expected to do them at home and have the solutions ready by the due date. Some quizzes were in-class — students were given time in class to do the quizzes, but they still got the quizzes in advance so they still had the option of attempting the questions before class.
• Discussion allowed or not: Some quizzes didn’t allow discussion. Some allowed discussion for all questions. Some allowed discussion for some questions. Students were free to use the notes or text for all quizzes.

Role of MCQ class quizzes in helping students stay synced

I believe that MCQs, with the particular design elements I chose, are a reasonable solution to the challenge of accomplishing the following objectives:

• Maintain a moderate amount of pressure and incentive for students to attend class.
• Make sure that if a student had to miss a class (or multiple classes) due to sickness or other reasons, he/she were not effectively penalized.
• In case that the student understood the material perfectly, make sure that the student did not need to attend class purely for bureaucratic reasons.

The way I believe this works is as follows: students feel some pressure to keep submitting quizzes. This creates a moderate incentive to attend class, because if they do not attend class, they need to email me the solutions, which can get awkward if they do this repeatedly with no legitimate reason to skip class. Also, since the quizzes test material closely related to what is covered in class, attending class can often save time with respect to attempting the quiz questions, so that the time gained by skipping classes is thereby reduced. On the other hand, if the student is sick or has some one-off reason for skipping class, there is no embarrassment or awkwardness in emailing the solutions.

Finally, a student who understands the material perfectly without attending class can download the quizzes online and email me the solutions, and end up saving a lot of time.

Role of MCQ class quizzes in providing feedback to the instructor and students

How well are students understanding what’s being taught? This can be hard for an instructor to judge, due to the illusion of transparency and double illusion of transparency. The MCQ-based class quizzes offer a relatively objective, independent, and quick way of keeping track of student progress.

Instructor burden

I spent effort on the MCQ class quizzes in the following respects:

• Designing and writing up the questions: This is a one-time cost for a particular course level since the quizzes can be reused. Estimated design time per question: 3-15 minutes, depending on the type of question.
• Writing up the solutions along with explanations: This again is a one-time cost for a particular course level. Estimated solution-writing time per question: 3-15 minutes, depending on the type of question.
• Printing, distributing quizzes, putting them up online: About 10-15 minutes per quiz (a quiz may have between 5 and 20 questions), but a repeated cost incurred each time the course is taught.
• Grading, preparing performance reviews, analysis: About 2 minutes per student plus about 1 minute per question (for question-specific analysis). A repeated cost incurred each time the course is taught.

The most substantial components of instructor cost are the one-time costs, therefore desigining MCQ class quizzes makes the most sense if you intend to teach the course repeatedly, and/or are teaching to a large number of students (because these costs do not scale with the number of students). I have also made my quizzes available for free, so if you’re teaching courses similar to mine and to a similar student body, you may feel free to reuse my quiz questions. Textbooks may also have MCQs or true/false questions that you can assign as quizzes instead of creating your own if you feel that the instructor cost of question design is significant, and/or you feel that the textbook adequately meets the needs of what you want to test your students on.

Why not have long-form quizzes?

Many of my colleagues as well as others I know, who agree about the benefits of using quizzes, default to using quizzes that differ from mine in the following respects:

• Lower frequency: Quizzes are once a week, in contrast to mine that are due almost every class (so about 2-3 times a week).
• In-class and requiring physical presence: Students are required to be physically present in class and take the quiz in class.
• Closed-book with no discussion: Students cannot consult their books, notes, or each other.
• Require long-form answers: Rather than MCQs, the questions require long-form answers.

This more standard approach to quizzes has the advantage that it requires a smaller one-time cost from instructors of question development. Designing good MCQs is harder than designing good long-form questions. The lower frequency and the fact that the quizzes occur in class means less fixed time devoted by students to the quizzes.

On the minus side:

• Instructors spend more time grading the quizzes (per quiz — if the frequency is lower, the overall grading time may still be less).
• It’s harder to do an easy combined analysis of the answers because they cannot be recorded easily in a spreadsheet or database.
• Students may feel more stress to attend class and to prepare for the quiz.
• The closed-book and long-form formats both severely limit the complexity of what can be tested. Students generally are better at understanding than expressing, and they are better at understanding with the help of aids. This allows for introducing them to more complex ideas.

What students said

Some anonymized snippets from student responses to the quizzes are below. Clearly, there’s a selection effect in terms of the students who share their feedback. Most of this feedback was received after the courses were over and I had no control over student grades.

• Snippet from handwritten feedback at the end of a course: “Thank you for making the extra effort to introduce us to interesting concepts without stressing us too much (see: quizzes)”
• Snippet from conversation with a student who was in my class and then took another math class with a different instructor: “While I hated the quizzes, it was actually a love-hate kind of deal, because while I hated doing them, they made me actually realize what we were learning. And now, when I don’t get anything in lecture and am stuck to learning things from the book directly, having quizzes would probably help a lot with learning what I’m supposed to learn, considering that homework is the only evaluation I have now of how much I understand the material.”