If you are new to active learning strategies, please [please, please, please] do not attempt to try everything all at once! It should be obvious, but that's a recipe for failure. That and improper implementation are the primary sources of frustrations such as "those methods don't work for me." Evidence now shows that research-validated interactive learning strategies can benefit ALL students in ALL classroom environments. But...the quality of our implementation is likely the most deterministic factor toward student achievement.

So choose one thing to focus on (Think-Pair-Share is typically considered the "gateway drug") and practice proper implementation until you are comfortable enough to add an additional component.

Remember to ask for help! You aren't the first to need assistance and you won't be the last. Even experienced implementers are always asking questions and seeking guidance. Don't go it alone!

Think-Pair-Share (or "Peer Instruction")


What is "think-pair-share" or "peer instruction"?

Think-Pair-Share (TPS) – or Peer Instruction (PI) as it's sometimes referred to – is an interactive learning technique that is research-validated in many disciplines, with Eric Mazur credited as its originator. Pose a conceptually rich multiple-choice question that students think about individually, vote on individually, then engage each other within specified parameters to reach a consensus that is usually correct.


How do I implement think-pair-share in my class?

How you implement think-pair-share (TPS) in your classroom is the key to success using the technique. Certainly you should tailor your implementation to fit your personality and course goals but there are some key points that shouldn't be compromised. There are also two rules for voting that are required to maintain the integrity of the technique: simultaneous and anonymous. My "crash course" accounting is below. The gory [but important] details are in the expandable sections below. There's also an older PDF handout here.

There is also the debate over whether you should use a "clicker" (sometimes also called a "classroom response system") or a "flashcard." There's also an Android app called TPSVote!! (and its older companion here) both available via sideloading. Whatever tool you use, make sure you are pedagogically motivated and not simply choosing based on the technology itself (pedagogy first, always). Most CAE collaborators – myself included – heavily favor the flashcards though some are very passionate about their clickers. See more on this debate in the "crash course" below. Also get my ABCD card here. I provide one per student on the first day of class and then post it on the course site for access at will.

A brave group of CAE collaborators uses the technique successfully – with flashcards, not clickers (!) – in Astro 101 "mega-courses" that have enrollments of 700 to 1400 (no, that's not a typo) and are taught in a performing arts hall. Check this out.

Read this first. (Seriously. Do NOT skip this part.) This is the Center for Astronomy Education's (CAE's) "how-to" guide written by workshop participant Amy Forestell Bartholomew who now helps coordinate CAE's Northeast Regional Teaching Exchange.

  1. I used to require textbook readings and Reading Questions prior to nearly every class period. Then in class I'd lecture to set up the tutorial. Now I'm dumping the textbook completely in "Astro 101," using only the Lecture Tutorial (LT) workbook, and flipping my class. So this first item will soon change to be something like "Have the students watch the appropriate mini-lecture video beforehand and complete the pre-class lecture questions."

In my "Astro 101" class the tests are all multiple choice questions (my other courses also make extensive use of multiple choice questions and the techniques found throughout this website). Therefore it is critically important that I give my students multiple opportunities to practice with the types of questions that they will experience on exams...like TPS questions. Similarly, since I make extensive use of LTs, I use TPS to set the stage for those exercises.
  1. When I'm highlighting the importance of our TPS activity as practice for the tests, I'll wrap up the background presentation (see #1 above) with something like this, "Cards! Let's see if you're ready for test day. This is an example of something you could encounter on your test."
    • My students know they need to have their ABCD cards out and ready to go when they walk into the classroom. When I yell "Cards!" that's their signal to grab it and get ready to vote in a few seconds because a TPS question is about to go up on the screen.
  2. When I'm using TPS to emphasize the importance of doing the LT I first start like in letter "a" above. After success with that first question, I move on to more complex ones, intentionally trying to get to one that most won't get correct even though it's completely appropriate and within the scope of what they've already covered. This highlights the need to work through the LT (or whatever other materials I've got in the queue) because they clearly aren't yet performing at the expected level and need some additional guidance.
    Note that if this is my motivation, I NEVER do the "pair-share" part of this challenging question that has them stumped. The point is to leave them hanging such that the LT helps them work through those issues and then we'll revisit TPS after completing the tutorial. That's where I'll get them to another question like that and then proceed with the "pair-share" parts below.

  1. It should be quiet as everyone must silently read the question, thinking about it on their own.
    Since the students first vote individually, it is critical that they not be influenced by others. I remind my students that this is a free chance to be honest with themselves and see what they really do and do not understand before being graded (exams, homework, etc.). I do not grade any aspect of TPS because I don't want them to give an answer just for the sake of getting points. It's part of my job to help them see and experience the importance of participating legitimately for learning.
  2. You do the same.
    Focus your attention on the screen (NOT them!), read it slowly to yourself just as they are – as if it's the first time you've ever encountered it – then silently go through each step in the reasoning process necessary to arrive at the correct answer, using your hands to "model" anything you need to (just as they should be doing). This accomplishes two goals.
    1. It validates the experience for the students by showing them that this is a real question that even you have to take the time to read and work through. Gesturing with your hands and modeling the situation to yourself (if that's a realistic method to use for that particular question) is a tool the students don't realize is available to them! They need to see that this physical act of modeling is a legitimate way to organize and process information.
    2. It ensures you give them an authentic experience by not reading the question to them. You should NEVER read the question aloud to them! Reading the question aloud is problematic for several reasons, most notably (1) it wastes valuable class time since they're just going to have to read it again to themselves when you get finished reading it and (2) it misrepresents the situation and can give away clues to the answer. It is misrepresentative because you won't be reading the questions aloud to them under any other circumstances, e.g. on test day, while they're practicing at home on their own, etc. And despite your best efforts to maintain neutrality, your vocal inflections and body language inadvertently give away clues to the answer.

If you've gauged your time properly in step 3 above you will probably begin to hear the rustling of cards right about the time you are finished "solving" the question yourself. Even if you don't notice the rustling of cards, if you've really tried to read and think like an "Astro 101" student you're probably on track with the time allotted. Turn to the class and ask "Does anyone need more time?" I suggest using these specific words. That is the precise question you need the answer to. If even one student signals yes, give them more time: I typically nod once affirmatively and turn back to the screen, devoting my attention there again while I slowly count to ten in my head. This shows them how important it is for them to really try and give a legitimate answer. (But only do this once!)

Say something like "Prepare your votes." Not everyone will have already done this so it's their signal to fold the card to the letter they want to show. (And don't ask them again if they need more time! They had their chance. See #4 above.)
  1. Voting A, B, C, or D means simply folding the card so that you show only the appropriate letter. Option "E" can be selected by folding it so that the back (only white) is displayed.
  2. There is a sixth option with a card that isn't available with a clicker: holding all four letters up at once, i.e. the paper is not folded. This is a legitimate vote in my class and it means they are so lost or confused that they can't even force themselves to hazard a guess. I do tell them that this option is not to be used as a "cop-out" and is truly a last resort since it's to their benefit to force themselves to commit to an answer even if they think it might be incorrect – that's important information to know!
  3. In the beginning when I am "training" them in the technique I tell my students to be a little stealthy about it. "It's anonymous, remember? You did a lot of work to determine the answer so you don't want anyone mooching off of you, right?" After a few class periods of implementation they of course get the hand of it and know that "prepare your votes" also means to keep it a secret (this is part of the anonymous rule).

To ensure the simultaneous rule, say something like "show your votes on the count of three" and count them off. On "three" they should be holding their votes underneath their faces and right up against their chests. This also ensures that the votes are anonymous with respect to the other students; only the individual student and the instructor know what answer the student has chosen. Holding the card in any other position means it's now possible for other students to see or figure out what letter is being shown.

Let the percentage of correct answers to determine your course of action. Don't be afraid of this part: your brain is much better at estimating the percentages than you think!
  1. If 80% or more got the question right then it's enough to move on.
  2. If less than 80% of the class gets the question correct, it's time for a discussion – the "pair-share" part.

Use the percentage of correct answers to determine your course of action.
  1. If 80% or more got the question right then it's enough to move on.
    1. You do always need to validate their correct response but resist the urge to simply tell them the answer yourself. It's much more pedagogically valuable if it comes from them. For example, I typically say something like, "That's excellent! The color of the right answer is..." and they are trained to yell out what color they voted. This is much less confusing than trying to hear the letter, all but one of which sound alike. If I don't get the level of confidence in the yell that I know I saw in the votes I'll make them do it again: "That wasn't very convincing. Let's try that again. The color of the right answer is..." and they will typically step it up. (Besides, who doesn't enjoy permission to yell during class?)
    2. This is a powerful moment for those who didn't get it right. When their incorrect "blue" or "green" responses are drowned out by a sea of people yelling "RED!" they just saw that not only was the question a fair one to ask but that nearly the entire class got it right...except for them. Those who got it right also become conditioned that you, as the instructor, are not the source of answers: they can do this, too, and need to learn to trust in their thought processes since that's what they'll ultimately need to be successful on the tests. This is, of course, why it is also critical to always validate the correct response and not leave them wondering. Repeat the color and/or letter of the correct answer such that all can hear.
    3. There are numerous other ways to debrief a correct answer. If you want to quickly reinforce a concept after they've called out the right answer, don't just launch into another lecture. Have them tell you the key pieces of reasoning such that they put it together themselves. This helps model the reasoning steps and highlights the understanding that giving answers isn't the way to be successful.
  2. If less than 80% of the class gets the question correct, it's time for a discussion – the "pair-share" part.
    1. First, do not reveal anything about the distribution of votes! In doing so, you bias their discussion and might give away the answer. There would be little, if any, deep engagement required on their parts to get it correct which leads them to not engage deeply enough to know why that is the answer. They just know it's correct so they're done thinking about it.
    2. If less than 50% of the class is correct, don't assume it's because they don't understand. You don't know that to be the case. Couldn't some of them have just misread the question? Read "west" when it actually says "east"? Somehow thought that arrow pointed up and not down? There's nothing wrong with letting a "rainbow" talk it out. They'll vote again afterwards and you'll take it from there.
    3. While they're still holding up their votes say, "When I say 'go' you have 60 seconds to convince your neighbors that you're right. Just because you have the same answer doesn't necessarily mean you're right so you'd better explain your reasoning. You have 60 seconds. Go!"
      • Be clear that their purpose is to convince each other. You don't want them to just "have a discussion": "Hey man I got 'A'; what did you get?" "I got 'B'." "Okay, cool." You need them to do precisely what you said: convince their neighbors that they are right by explaining their reasoning. It forces them to own up to the answer they chose, defend it, and try to debunk others' different answers.
      • You also put everyone's vote in jeopardy...on purpose! That's the point: any given student could be right or s/he could be wrong. That's why you don't give away anything about the first vote's distribution. It stifles any attempt at genuine debate.
      • You've also created a sense of urgency with the time limit (60 seconds). There's no time to waste off topic. If they're slow to start, begin counting backwards out loud, "59...58...57..." That works like nothing else does. (Just like saying "go" is magic compared to other terms you might try there.)

As they're talking, circulate around the room listening closely. You may want to carry a notepad around since this is a really great opportunity to jot down ideas for new multiple-choice questions and distractors!

Also, the time limit is bogus. I almost always say 60 seconds regardless. Here's why. As I'm circulating and listening I can tell whether we're on track or not. If I hear the conversation dying down significantly and it's been less than 60 seconds, I'm going to shut it down early since I don't have class time to waste. If, after about 60 seconds, I hear that the conversation is still rich and on topic, I'll let them go until I'm satisfied with it. Either way, I give them a warning, "You have about ten more seconds," while I'm making my way back to the front of the room. Then, it's time to vote again.

Repeat steps 5 and 6 above to re-vote.

At this point, you're probably up to the 80+% level and can proceed as in steps 7a and 8a above. If not, there are still options available.
  1. Do you see any patterns in the distribution? Bimodal? All correct except for the group in the back left? Predominantly wrong because they fell for the distractor that means they still didn't notice the direction of that arrow? You could try the following.
    • "Did we all notice that this arrow points down and not up?" You might then hear a chorus of "Oh!" Time for another vote.
    • "Is there something about the question that I can clarify for you?" You might be able to dispel confusion to their satisfaction and vote again.
  2. Then again, you might realize there is actually a problem with the question. It happens, especially if you're piloting a new one you just wrote! Or perhaps this question is just too complex for them right now: you took too big of a leap in reasoning steps. Back up to a less complex question that tests only the core idea they seem to be struggling with then build up the additional reasoning steps.

Work by CAE collaborators indicates that using flashcards results in greater participation rates and more correct answers than clickers despite no changes to the methodology, i.e. simply exchanged a clicker for a flashcard. This is likely attributed to the more personal nature of, and social aspects embedded in, using a card (see the last paragraph of this paper, before the acknowledgments).
  1. Probably the biggest drawback to using clickers instead of cards is the inability to tie a vote to a specific individual in real time. Sure you can assign the same numbered clicker to the same student all term, but how long does it take you to see exactly who is selecting what answers? Can you tell if that group of incorrect responses is all folks sitting together in the same part of the room? Can you tell whether individuals made progress from a first to second vote? The real-time benefit of the cards here is unmatched.
  2. One plus for clickers is the ability to save the data. This sounds like a clincher and if you're conducting research using the numbers it very well may be. But be honest with yourself: how often are you really going to refer back to the data? Do you really need exact numbers? Or would jotting down a quick note about that question suffice? Have you seen the same patterns over and over such that you really need the numbers?
  3. Then there's the argument of using the clicker to assign participation credit. I NEVER grade any part of TPS and wouldn't ever advocate for it. While some do (even CAE collaborators), there's so far no correlation of credit with learning gains that we know of. Of course the students get really good at "gaming the system" when they know points are earned just by pressing a button on a remote control in anonymity. (There are even bizarre anecdotes about students "buzzing in" for absentee friends by taking the absentee students' clickers in addition to their own or by doing "drive bys" and clicking from their cars since the wireless range on some of them is pretty incredible!) Besides, there are tons of other ways to assign participation credit that span the gamut from those with real pedagogical value to some as easy to grade as looking at a clicker roster.
  4. What about the cost to everyone, especially the students? What about set-up and take-down time, techno-glitches, dead batteries, forgotten/lost clickers, etc.? How about the ability to insert new questions on the fly? Cards easily win all these rounds. Granted, some of the more user-friendly clicker systems are making it easier to insert questions on the spur-of-the-moment. However, the instructor still has to set up everything properly. Cards are always ready to go – even when you want to test a question from a tutorial or out of the chapter (or even make up a question off the top of your head and simply say it aloud [careful, though; see number 3.b.ii. above; it might be okay depending upon the purpose]). Even if a student forgets his/her card, s/he can "give me the finger" (or fingers: ☝, ✌, 👌, etc.) during that class period. Because it's harder to call out the color of the right answer when you voted with two fingers instead of "green," they usually get themselves a new card pretty quickly.

Regardless of the method of voting you use with TPS, remember: (1) pedagogy first and (2) maintain the integrity of the technique by practicing proper implementation and keeping with the rules of simultaneous and anonymous.

Where do I get think-pair-share materials?

If you are using TPS as your first venture into active learning strategies you may need to write some – but not all – of your own questions (see "Other" for info on writing good multiple-choice questions).

Alternatively, if you are also using astronomy Lecture Tutorials in your course know that there is a good selection of pre- and post-LT questions for each tutorial contained within the instructor's manual (you will have to register and get verified as an instructor on Pearson's site to access it). Similarly, the Center for Astronomy Education has exam banks available for download from their workshop materials page. A sample of some of the more complex questions, requiring several reasoning steps each, is available here.


Additional Resources

There are some published resources with questions and implementation guidance. Not all are research-validated and they have varying degrees of utility.
MiraCosta College Astronomy Program
tiny.cc

general site © Rica Sirbaugh French 2017-18
Many of the design elements, functionality, and materials presented herein are *not* mine and I claim no rights to them.

last edited:
current load: