Tag Archives: CGI

Children’s experiences with partitioning [TDI 3]

If you watch this video, you will see a pretty standard U.S. treatment of introductory fraction material.

PLEASE understand that this is not about Sal Khan or Khan Academy. What you see in that video is what happens in many, many elementary classrooms across the U.S. on any given day. It is what is written into our textbooks (pre-Common Core, of course—we’ll get to Common Core in week 6).

I did not have to work very hard to find additional examples to support my claim. Here is a tutorial on Sophia. Here is something from “Your destination for math education”. And here is a self-paced math tutorial at Syracuse.

I am not cherry picking straw men here.

To be clear, we introduce fractions with a part-whole model. A circle (or rectangle) represents the whole. We cut the circle (or rectangle) into some number of equal-sized pieces—that number is the denominator. We shade some of those pieces—that number is the numerator. That I am pointing this out surely makes some Decimal Institute attendees uncomfortable because how could it be any different? I’ll get to that in a moment. Stick with me here as I build a case pertaining to decimals.

If you believe that defining an abstract mathematical object and then operating on that object is the most powerful way to teach mathematics, then there is no logical objection to starting fraction instruction with decimals.

After all, children know something about our base-10 place value system by the time they get to third grade. They know something about the decimal point notation by then, too, as the result of money and (sometimes) measurement. (Oh, and calculators—don’t forget the calculators.)

So why not put all of that together and have tenths—the very heart of the territory to the right of the decimal point—be the first fractions they study? If you believe that children learn mathematics as a logical system that is little influenced by their everyday experience then there is no reason not to.

From a logical perspective, halves and tenths are the same sorts of objects. Tenths come along with a handy notation and so—from a logical perspective—are simpler than halves.

Indeed, it is much much easier to train children to get correct answers to decimal addition problems than it is to train them to get correct answers to fraction addition problems—even when the fraction addition problems have common denominators. (Sorry, no research link on this. Ask your nearest upper elementary or middle school teacher whether I am talking nonsense here.)

But we cannot fool ourselves into believing that ease of obtaining correct results has any correlation with grasping underlying concepts. Children can be trained to give correct answers without having any idea what the symbols they are operating on represent.

Take the video linked here, for example. (In it, I do a Khan Academy exercise using a purposely flawed way of thinking and score approximately 90%—I get an A without showing that I know anything useful.)

This leads us to the work of the Cognitively Guided Instruction (or CGI) research project from the University of Wisconsin. This project studied the ideas about addition and subtraction situations, and strategies for working them out, that children have before formal instruction begins.

It turns out that they know a lot.

Of particular importance is their finding that when teachers know how students are likely to think about addition and subtraction problems, and when teachers know the strategies students are likely to use, these teachers are more effective at teaching addition and subtraction.

In short, CGI demonstrated—for addition and subtraction—that better understanding the cognitive structure of addition and subtraction makes you a more effective teacher.

In the years since that first set of results, the team has extended their results to initial fraction ideas. In the book Extending Children’s Mathematics, they argue that the cognitive way into fractions with children is fair sharing.

That is, the ideas that children bring to school prior to formal instruction having to do with fractions are those that come from sharing things. Sharing cookies, cupcakes, couches and pears; children have cut or broken these things in half, considered whether the resulting pieces are equal in size, and decided whether the sharing is fair many times before they study fractions in school.

When you do start with fair sharing, children’s ideas about how to do this follow a predictable path. Halving and halving again are common early ideas even when sharing among three or five people. Similarly, children share incompletely early on. When they need to share one cookie among 3 people, they will suggest cutting into 4 pieces and saving the fourth for later.

This more recent CGI research demonstrates that paying careful simultaneous attention to (1) the number of things being shared, and (2) the number of people doing the sharing is a late-developing and sophisticated skill that comes as an end product of instruction.

You can see this in a conversation I had with my children over the weekend (written up in full on Talking Math with Your Kids).

In that conversation, we had 2 pears to share among 3 of us (real pears, not textbook pears). Griffin (9 years old) suggested cutting them into thirds, but then got distracted by the campfire before correctly naming the amount we would each get. Tabitha (6 years old) worked with me to half and half again. Only once we had a single remaining piece right there in front of us did she suggest cutting that piece into 3 pieces.

The concrete conversation created a need for thirds. But thirds only occurred to her once that need existed. As long as we had whole pears or halves of pears, we could keep cutting in half.

Here was the end result of that sharing.

Photo Oct 12, 2 13 41 PM

Now back to decimals.

The CGI fraction work constitutes persuasive evidence that not all fractions are cognitively equivalent. While starting the study of fractions with tenths makes sense from a logical perspective, CGI demonstrates that children do not learn from logical first principles.

They learn by considering their experience.

Children have lots of experience with halves. We might expect thirds to be just as obvious to children as halves are, but it isn’t true.

So let’s take seriously the idea that experience in the world has an effect on how children learn. And let’s accept that this fact should have an effect on curriculum design.

Then if you still want to teach decimals before fractions, you would have a responsibility to demonstrate that children have anywhere near the real-world experience with tenths that they do with halves and thirds.

When we discussed on Twitter recently children’s real-world experience with tenths, we came up with:

  • money (where the connection to fractions is weak, see also week 2’s discussion on Canvas),
  • pizzas (about which I am skeptical, see next week’s interlude),
  • metric measurements, and
  • not much else.

In comparison to the tremendous amount of work children have done with halves and halves of halves (and halves of those), how can tenths be the first fraction they study in school?


To summarize, I am arguing:

  • That part-whole fraction work makes logical sense to experienced fraction learners,
  • That children do not learn fractions by logical progression from definitions, but by connecting to their experiences with situations in which fractions arise in their everyday lives,
  • That we have research evidence for this latter claim,
  • That the truth of this claim should have implications for how we teach decimals to children, since their experiences with tenths are much less robust than their experiences with simpler fractions, and that chief among these implications is…
  • That we ought to reserve serious decimal work until kids have developed the major fraction ideas about partitioning, repartitioning and naming the units that result.

Guess the temperature

Griffin and I play a little game called Guess the Temperature. It goes about how you would expect. We step outside on the way to his bus. I ask him to guess the temperature. If I don’t already know, I get to guess after he does. If I do already know, I don’t cheat; we just remark on how close his guess was.

In Minnesota, this means we get to study integers.

Me: Griff, guess the temperature.

Griffin (eight years old): Two below zero.

Me: It’s three degrees above.

G: So I was off.

Me: Not by much, though. How much were you off by?

G: [muttering to himself, then loudly] Five degrees!

Me: How did you know that?

G: It’s two degrees up to zero, then three more.

Let’s pause for a moment here. You know how I just won’t shut up about CGI (Cognitively Guided Instruction)? It’s because they’re right. Children know mathematics before it is formally taught.

Consider the grade 6 (for 11-year olds) Common Core Standard 6.NS.C.5

Understand that positive and negative numbers are used together to describe quantities having opposite directions or values (e.g., temperature above/below zero, elevation above/below sea level, credits/debits, positive/negative electric charge); use positive and negative numbers to represent quantities in real-world contexts, explaining the meaning of 0 in each situation.

Griff pretty much has this nailed down and is making progress on grade 7. But no one has formally taught him how to subtract integers. He reasons his way through a problem by making sense of the relationships in the context. He can find 3-(-2) without knowing keep-change-change.

But it’s not just Griffin. CGI demonstrated that children—all children—develop mathematical models of their worlds that precede instruction, and that instruction sensitive to these mathematical models is better than instruction that ignores them.

Back to our conversation.

Me: So what if it 10 degrees out, and you guessed 3?

G: [quickly] I’d be seven off.

Me: Right. How do you know that?

G: Ten minus three is seven.

Me: Nice. Subtraction. Do you know that you can always express the difference between your guess and the actual temperature with subtraction?

So in that last example, you subtracted your guess from the actual temperature. You could do that with your real guess today.

So three minus negative 2 is five.

G: [silent]

By this time we were nearing the bus stop. I had offered this tidbit as an intellectual nugget to chew on, rather than a lesson I expected him to absorb. But that is what it means to have instruction be sensitive to children’s mathematical models.

Another cool thing…

In our discussion the other day of whether the difference between van Hiele levels 0 and 1 is the use of official math vocabulary terms, or whether there is something more there, a student asked this (I am paraphrasing):

Is this like the CGI research, where children know things about addition and subtraction before they are taught in school?

Here’s what I love about this question:

  1. It demonstrates the power of having these students for a second semester;
  2. It demonstrates that students will—over time—make these ideas their own; if we keep at it, and build cases for the importance of the ideas in our courses, students will play with these ideas, and they will look for opportunities to apply and connect them; and
  3. It made me stop and think. I’m still not 100% sure whether it’s like that or not. I have something new to think about.

Salami-slice math

I was making lunch for Tabitha (who had just turned 5) one day. It was one of her favorites-salami quesadilla (house specialty). She was impatient and hungry. She asked for some salami while she waited.

Me: How many slices do you want?

Tabitha: Four.

Me: I’m sorry to say, we only have three.

T: Oh. That’s one less than I wanted.

Me: That’s right. Do you want them?

T: Yes.

Me: (Giving Tabitha the remaining salami slices) So if three slices is one less than four, what would two less be?

T: Two.

Me: And what would three less be?

T: (Thinking) One.

Me: Nice. How did you know that?

T: I don’t know.

Me: What would four less than four be?

T: (Thinking) Zero!

Me: Nice.

T: Ask me another.

Me: What would five less than four be?

T: (Thinking) Zero?

Me: Interesting.

T: Do a different “less”. Like less than five or something.

There are two important strategies here. The first is turning Tabitha’s request for salami into a situation that involves numbers. She asked for salami to eat while waited for the main part of her lunch. I could have just given her a few pieces of salami, but that’s a lost opportunity. It’s an easy move to ask her how many slices she wants, and then to compare that to the number of slices we actually have.

The second important strategy is the What if? questioning style here. This is where the abstraction happens. When I ask her What would two less be? I am offering here the opportunity to think in terms of salami slices, or to just think about numbers. She has already imagined the slices so they are available as a tool. Or she can count numbers in her head.

Note Tabitha’s comfort with zero as a number. When I ask her what four less than four is, she comfortably answers zero. Notably, she has to think about it. She hasn’t developed the rule that any number minus itself is zero, although this conversation could certainly go that direction by asking about three less than three salami slices and five less than five salami slices, etc.

Tabitha is five years old. There is plenty of time for her to learn about negative numbers. I did her no harm asking What is five less than four? I didn’t expect her to be able to say negative one, but I was curious about what she would say. Remember that being interested in children’s ideas is a key to talking math with them.

It matters, though, that she didn’t just reverse the numbers and say “One”. The numbers 5 and 4 have very specific meanings for her, and she uses those meanings the best she can.

Kindergarten addition strategies

A few months back, Tabitha (5) wanted me to ask her some math class questions. This led to some disagreement between the two of us on what constituted a math class question.

On our way back from our fall camping trip last weekend, Tabitha announced that she knows what 8+8 is.

Me: Ok. What is 8+8?

Tabitha: [longish thinking pause, eyes gazing up and to the right] 16!

Me: Wow. I could really see you thinking that through. How did you know that?

Tabitha: I counted.

Me: OK. Good. But how did you keep track of how much to count?

Tabitha: huh?

Me: How did you know when to stop counting?

Tabitha: When I got to 8. That was 16.

Me: Right. But how did you know when you had counted eight times?

Tabitha: I just counted to 8. That was 16.

Me: Oh! I know! Was it like this? Eight…nine (that’s one), ten (that’s two)…like that until you got to sixteen (that’s eight)?

Tabitha: Yes.

Me: How did you know to do that? Did you learn to do that in Kindergarten? Or was it your own idea?

Tabitha: [long pause] OK. I admit it; it was my idea.

I have written before about how sticking with this stuff pays dividends eventually. Tabitha announced that she knows what 8+8 is, but what she really meant was that she knows how to find 8+8. Tabitha knows that talking math is not about telling answers, but about actively thinking things through. That doesn’t come for free in classrooms, and it doesn’t come for free at home. It is a result of lots of prep work and lots of encouragement and lots of conversations.

Returning to the theme of the knowledge required for me to have this conversation with Tabitha, though, I needed to know something about possible strategies, and I needed to have thought some of those strategies through for myself-by trying them out.

If the sum in question had been 8+2, I wouldn’t have questioned her claim that “I counted”. You can subitize two, so you can know that you have counted twice without having a separate system for keeping track.

But counting 8 times without having a way to ensure that you have done so? This seems pretty much impossible. As an adult, I know to stop at 16, but that’s only because I already know the sum. As I count 9, 10, 11, 12, I am conscious that the only way I know I have counted four times is because I know that 8+4 is 12. I’m not able to keep track of the number of counts past three.

I had watched her fingers; no motion there.

So there had to be something else going on. I was only able to press because I knew what she knew (how to count) and what she didn’t know (addition facts) and some possible ways that this knowledge could be used to find the sum of 8 and 8.

The more I talk with these kids, the more impressed I am with the work of CGI. Much of what I know in this area is the result of studying their work.

UPDATE: The original CGI link in that last paragraph went to a tag search of CGI on my blog. That was confusing. It has been replaced by a particular post. Here’s the original link. And here’s a link to CGI on Wikipedia.