I was recently discussing the use of a Think, Pair, Share (TPS) with a colleague. It does what it says on the tin: you pose a question or a text and have students first think about it, then pair up and share their thoughts. It’s pretty common, and I’ve seen lots of resources with a TPS built in and lots of people talking about it online.
My colleague was planning on using a TPS on the introduction of a new topic, but one which students might know a little bit about first. According to Wikipedia, this is how TPS’s inventor Frank Lyman intended it. I don’t really do this in my lessons, and though I do see value in students collaborating and working together, I wanted to flesh out my thoughts a little and explain why it might be better to rethink this particular strategy.
Let’s assume that the thing to be TPS-ed about is a question, like “what is photosynthesis?” In every pair you have two students involved, each student with four potential characteristics:
- They know what photosynthesis is (K)
- They sort of know what photosynthesis is (k)
- They don’t know what photosynthesis is and know they don’t know it (N)
- They think they know what photosynthesis is, but are wrong
As a table of possibilities you get:
To my mind, “good” outcomes are ones either where students have the opportunity to recall something they already know well (KK) or where one student is a K and can teach a k or an N (though even in this case, one could argue that it would be “better” if the teacher did that teaching). Either way, that leaves us with positive (green) and negative (red) outcomes:
This isn’t a pretty picture – a lot of these outcomes are not only not positive, but actually harmful. Students who are W obviously result in the worst possible outcomes, as they will actively teach students who are k or N the wrong thing, and could result in over-riding students who were K before, meaning that students who started the lesson with the right knowledge could leave it with the wrong knowledge. And even if you correct the “wrong knowledge” afterwards with your teaching, we know full well that students will often remember the wrong thing from a lesson, and you have no guarantee they will come away with the “right” answer and not the “wrong” one.
In the red options that don’t have a W we have kk, Nk, and NN. If students don’t really know something, and are asked to discuss it, you can fairly ask whether this is a good use of their time. Two kk students who don’t know much about photosynthesis talking about photosynthesis are hardly likely to magically come up with the correct definition, word equation or explanation as to where the reactants come from or the products go to – and how much more so if they are Nk or NN.
There also might be a social justice/equity answer here. If the only students who are benefiting from your task are ones who already know things about what you are teaching, then they are being unfairly favoured. You end up with students who come from a certain background as getting something from your activity, and ones from another background as just falling further behind. Any time your activity relies on students having some knowledge of a topic from outside school you run a very serious risk of further disadvantaging the already disadvantaged.
There’s probably room to disagree between subjects here as well. In science, there are clearly wrong answers. In other subjects, there can be more wiggle room and areas that are more up for debate, and I wouldn’t want to talk about subjects I don’t teach. But if we took something that is up for debate in science like organ donation or stem cell research or whatever, I still don’t see how students will be able to have complex thoughts about this before a teacher has taught them about it.
In sum, I don’t really see the point. Again, I really do think there is a time and a place for students to work in pairs, but for me that time and place is when drilling knowledge already learned and holding each other to account for use of correct terminology and keywords. But if it’s about them exploring something new, it doesn’t seem like a particularly effective strategy
No doubt, there will be someone who takes me to task on Twitter for saying I haven’t understood TPS, or they do it really well, or it works really well for them, or something something active learning. There’s probably a paper somewhere where some researcher observed students doing a TPS and reported higher levels of engagement. Fine. If you think that none of the points I raised above apply to your classroom that’s cool. But if you were thinking of doing it just because it’s the done thing or it’s what an observer expects to see or because of some semi-formed idea of social constructivism, please do consider alternatives first.
March 15, 2019 at 6:06 pm
Hi Adam. It’s great to get into the detail on these things. I know what you mean in terms of the ‘blind leading the blind’ issue. However, I find that TPS is hugely useful in the flow of a lesson to detect those k N and W students. Often I see teachers create bottlenecks in questioning when it all has to go through them.. and then it’s not always efficient to set a test ….so a quick time-bound bout of TPS with a certain goal (such as naming the reactants in photosynthesis,, or list the five Group 1 metals in order of reactivity -) followed by some cold calling, can help flush them out. Arguably miniwhiteboards would do this too– but TPS allows students to walk through something (explain the water cycle), in a little bubble of safety. I see it work well very often even with some of those red pairings: the TPS is the means by which the teacher finds out what they didn’t know, more quickly than in many other methods. In fact I’d say that in the absence of TPS where every student is engaged in some form of dialogue it’s often more likely that teachers make overly-optimistic assumptions about the level of understanding around the class because more generative recall is happening and being opened up to scrutiny.
The risk is that they’ll inadvertently remember all the wrongness being explored… so you definitely need TPS to link to a verification process: Did you get it right? They need to see right answers or secure information to check their own against. If everyone is left thinking that whatever they shared was ok, that’s no use. I saw this recently when a teacher asked for a definition of ‘pollution’.. sampled a few responses from pairs.. and basically approved them all even though some were quite loose and problematic. That’s risky. So… definitely the way that you do it matters a lot.
LikeLike
March 16, 2019 at 8:39 am
For reasons you mentioned, it is probably not the right move to use TPS for a retrieval questions. For me, the function of TPS is as a thought and thought accountability tool. In other words, if I am teaching a poem My Last Duchess and I want students to try and puzzle out what it means that the duchess is described as , “Looking as if she were alive.” Now, it’s unlikely that any student is not going to know what those words mean (with the possible exception of were).
The value in a TPS (or, actually I’d probably use a quick-write before the PS) is that the “think ratio” is higher if we ask students to work out what this could mean. It’s a problem that students can solve, and in solving it they will think about content, and I thinking they will learn. Memory is the residue of thought.
LikeLike