Are You Using Think-Pair-Share Correctly?

Do you use think-pair-share in your classes? Are you doing it right?

Recently, I (virtually) saw Eric Mazur speak on Peer Instruction, his teaching technique that incorporates a variation of traditional think-pair-share (Mazur, 1997). This reminded me of key aspects of think-pair-share that I had either forgotten or haven’t always implemented. Thinking back to my experience as a student, I think instructors often miss key aspects of think-pair-share, reducing its effectiveness.

You may be thinking, “Really? Isn’t the process fully explained in the name?”

Sort of, but there is surprising nuance involved in well-executed think-pair-share, especially the highly effective version used in Peer Instruction.


In the traditional conception of think-pair-share (originally proposed by Frank Lyman (1981), though I’ve never found a copy of that piece):

  1. The instructor poses a question.
  2. Students ponder the question individually.
  3. Students share their thoughts or answer with a classmate.
  4. Optionally, some pairs may report their answer to the whole class.

Peer Instruction

In Peer Instruction, the steps are similar, but the approach is more regimented (Crouch & Mazur, 2001):

  1. Question. Instructor poses a question, based on an assigned reading or recently presented information.
  2. Think. Students consider the question individually.
  3. Poll. Students commit to an answer. This can be done with a classroom response system (e.g., clickers) or a low-tech approach like holding up a colored index card.
  4. Discuss. Students discuss their chosen answers in pairs or small groups.
  5. Re-poll. Students again individually choose an answer to the question.
  6. Explain. Instructor explains the correct answer.

The Questions

Think-pair-share works best when the questions are very carefully crafted. Mazur calls his questions ConcepTests (conceptual tests), and they require higher-order thinking and/or application of recently learned ideas. These are multiple choice questions with plausible distractors that often target students’ misconceptions. Mazur suggests question difficulty such that ~50% of students get the correct answers on the first poll.

Where can implementation of think-pair-share go astray?

1. Skipping the “Think”

I have done this. I toss out a question to students (often spontaneously) and ask them to discuss it in pairs or small groups. This isn’t a complete failure, but if students immediately start discussing in small groups, it is easy for some students to remain cognitively disengaged. They can simply listen to a classmate’s answer without ever thoughtfully considering the question themselves.

2. Not requiring students to commit to an answer

Similarly, committing to an answer increases the likelihood of engagement, and students also become more emotionally invested in the result (“Did I get it right?!”).

3. Uninteresting question

If a question isn’t relevant, intriguing, puzzling, or exciting, it will feel like busywork or just a waste of time.

4. Question too easy or too hard

If a question is too hard, say only 10% get the correct answer, then peer discussion is unlikely to improve students’ understanding. Likewise, with an easy question, students won’t benefit from discussing something they all agree on. Though the target is 50% correct answers, Mazur suggests that a range of 30% to 70% is suitable (Lasry et al., 2008), though an instructor can make adjustments on the fly if results are outside this range. If fewer than ~30% get the question correct initially, the instructor can revisit the concept and have students choose an answer again before proceeding with the normal approach. If more than ~70% of students are correct, the peer discussion step can simply be skipped (Lasry et al., 2008).

Adapted From Lasry et al. (2018)

Other Benefits of Peer Instruction / Think-Pair-Share

I’ve written a few times about effective learning strategies, especially retrieval practice and spaced practice, but there are others. In self-explanation, a student explains the reasoning behind the steps they took to solve a problem (Berry, 1983). In their review of learning strategies, Dunlosky et al. (2013) rated self-explanation as having medium utility (only retrieval practice and spaced practice were rated high utility). It should be immediately apparent that if students in the classroom are thoughtfully discussing and justifying their answers with peers, they are likely engaging in some variation of self-explanation.

In introductory college physics courses, peer instruction has improved conceptual reasoning and quantitative problem solving (Crouch & Mazur, 2001; Fagen et al., 2002) and reduced rates of student attrition (Lasry et al., 2008). If think-pair-share is graded only for participation, then as Ken Bain (2021) observes, “The whole process employs formative assessment for as long as possible, giving people plenty of opportunity to try, fail, get feedback, and try again before anyone makes a judgment about their progress.”

While applications of think-pair-share are often considered in STEM fields, this can also be used in the humanities. I’ve referred to “correct answers” for simplicity, but in Mazur’s talk, he used the term “preferred answer” to clarify that the approach can still be used even when there may not be a single, objectively correct answer.

Effective Use of Think-Pair-Share

In summary, think pair share is most effective if you follow a few key practices:

  • Most critically: Pose thoughtful, meaningful, and intriguing questions of moderate difficulty. Target higher-order learning (not simply recall) and craft the incorrect answers around student misconceptions. Keep notes on how well certain questions work and revise them over time.
  • Give students enough time to thoughtfully answer the question on their own.
  • Have students commit to an answer before discussing with peers. Use a classroom response system if available or simply have students hold up the number of fingers corresponding to their answer, e.g., 1 for A, 2 for B, and so forth. (Mazur has students hold fingers against their chest, so they are visible to the instructor but not obviously visible to peers).
  • Don’t penalize students for incorrect answers! Let them engage with the concepts and focus on their reasoning and understanding without the stress of grades.

If you already use think-pair-share, think about whether you could improve your implementation. If you don’t use of it, it is a relatively straightforward approach to improving student engagement in the classroom.


Think-pair-share is a relatively straightforward, highly effective approach to improving student engagement, even in a large lecture setting. There is some hidden nuance though. For best results, implementation needs to be quite deliberate, and questions must be thoughtfully crafted. If you already use think-pair-share, take some time to reflect on your implementation. Are you meeting all the key requirements? Are your questions exciting? Do they engage your students in higher-order reasoning? Where could you make improvements? If you haven’t used this approach before, it’s a good one to add to your repertoire.


Bain, K. (2021). Super Courses: The Future of Teaching and Learning. Princeton University Press.

Berry, D. C. (1983). Metacognitive Experience and Transfer of Logical Reasoning. The Quarterly Journal of Experimental Psychology Section A, 35(1), 39–49.

Crouch, C. H., & Mazur, E. (2001). Peer Instruction: Ten years of experience and results. American Journal of Physics, 69(9), 970–977.

Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving Students’ Learning With Effective Learning Techniques Promising Directions From Cognitive and Educational Psychology. Psychological Science in the Public Interest, 14(1), 4–58.

Fagen, A. P., Crouch, C. H., & Mazur, E. (2002). Peer Instruction: Results from a Range of Classrooms. The Physics Teacher, 40(4), 206–209.

Lasry, N., Mazur, E., & Watkins, J. (2008). Peer instruction: From Harvard to the two-year college. American Journal of Physics, 76(11), 1066–1069.

Lyman, F. (1981). The Responsive Classroom Discussion. In A. S. Anderson (Ed.), Mainstreaming Digest (pp. 109–113). University of Maryland College of Education.

Mazur, E. (1997). Peer Instruction: A User’s Manual. Prentice Hall.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.