# Tag Archives: bar models

## Teaching Ratio

Novices and experts see problems differently.  Whereas a novice sees superficial features, an expert notices deeper underlying patterns, discarding the often irrelevant and distracting contextual information.  Here’s an example:

Filed under Maths

## Acronyms like RUCSAC prevent children from thinking mathematically – we need a different approach

I’ve got a thing about success criteria. Very often, the line between what we want children to learn to do and the task that we ask them to carry out is blurred. The gap is perhaps most stark when it comes to problem-solving in maths.

In many classrooms the “read, underline, calculate, solve, answer, check” (RUCSAC) acronym, or something similar, will be plastered on the wall and used as success criteria for problem-solving.

However, I’d argue that RUCSAC does not present a valid set of criteria for such an important part of maths; rather it prevents children from learning to think mathematically. Here’s why…

Read the rest of the article here.

1 Comment

Filed under CPD, Maths

## A place for everything and everything in its place

Place value is very often one of the first units of work for maths in most year groups and is absolutely fundamental to a good understanding of number.  By getting this right and giving children the opportunity for deep conceptual understanding, we can lay solid foundations for the year.

For the purpose of this blog I’m going to assume that children can count reliably and read and write numbers without error. If these things are not yet developed to the appropriate standard then targeted intervention needs to happen without the child missing out on good modelling and explanations of place value.

Children need plenty of practice constructing and deconstructing numbers, first using concrete manipulatives like base ten blocks or Numicon.  This is to show that 10 ones is equivalent to 1 ten etc.  While they’re making these numbers they should be supported to talk articulately about what they are doing, perhaps with speaking frames: ‘This number is 45.  It has 4 tens and 5 ones.  45 is equal to 40 add 5.’

Read the rest of the article on the Rising Stars Blog.

Filed under Maths

Moving schools and with more than an eye on headship is sure to get you reflecting.  The following posts are what I think about various things, in no particular order.  Previous posts were about displays and learning.  Next up – maths.

Any unit of work should be planned with end points in mind.  Teachers should start with the relevant National Curriculum statements but more importantly, the types of questions that children will be expected to be able to solve.  Teachers in Y2 and Y6, with their experience of end of key stage testing, may well have internalised the type of questions that would be appropriate for children to solve but resources like the Rising Stars Assessment Bank are invaluable for teachers.  Our expectations of what children will learn are vital. All children are capable of achieving age related expectations given the right support and sufficient time. If we begin a unit with lowered expectations for some children based on perceived ability then we are failing them.

Once the expectations of what children will be able to do at the end of the unit are clarified, teachers should then plan backwards, thinking carefully about what children will need to know and be able to do in order to solve those problems as well as figuring out a sensible conceptual sequence of those things.  For me, these include fluency with number, learning the underlying structure of the problems to be solved, the deliberate teaching of mathematical language and opportunities to reason.  The more I think about this, the more I’m settling on a sequence of development that units of work should be structured around, key parts of the sequence are developed and consolidated over time:

This model may not work for all topics but I’d suggest that units of work start with pure number.  Contexts can be stripped away to reduce load on working memory and children can get on with learning and practising fluency so that as soon as possible, they are able to recall necessary factual knowledge and manipulate numbers in calculations with little mental effort.  This is by no means rote learning – it should be carefully thought out so that children develop sound conceptual understanding, starting with concrete representations, progressing to more efficient pictorial representations and then on to even more efficient abstract representations.  It is important that teachers remember that the abstract representations are not the maths itself, merely the most efficient way of recording or communicating the thinking.

Fluency

If children can recall number facts and other basic mathematical knowledge within a few seconds, if children can calculate reliably without expending too much mental effort, and if children can recall varied mathematical knowledge, switching between topics, then they will be considerably more able to commit precious working memory capacity to problems that require deeper thinking.  It is for this reason that fluency must come first and continue to be practised in order for that recall to become increasingly accurate and efficient.  Flash cards are very useful here – they provide the opportunity for self-testing and, with a little training, can help the child to become more aware of the what they know and do not know, enabling them to focus their own study.  Teachers figure out what basics are required and deliberately teach those basics if they are not sufficiently internalised already.  If some children are already fluent, they can work on speed and efficiency, for children may well have fluent but inflexible strategies.

Underlying structure

Stories can be said to be a variation or combination of just seven basic plots and expert writers have a sound knowledge of these, enabling them to see stories at a deeper level.  I’d argue that there is a significant similarity in maths; that there are five basic problem structures.  These five structures are aspects of either additive or multiplicative reasoning and are classified based on what is unknown in a problem:

Knowing the structures is not sufficient.  Children must be able to identify the underlying structure from a given problem.  This is no mean feat so sticking to just number problems and avoiding distracting contexts for the time being is important.  Take a problem like this:

Hattie and Yates, in Visible Learning and the Science of How We Learn, said that experts see and represent problems at a deeper level.  A novice will only see the surface features in this problem: two numbers and an addition sign.  Consequently, they’ll solve this by adding the two numbers together.  Of course, this is a mistake.  An expert, on the other hand, will know that the whole is made up of parts and that you add the parts together to make the whole.  They may even ignore the numbers at first and read it as part + part = whole, realising that it is a problem where one of the parts is unknown.  They may see it or draw it like this:

They’ll use their knowledge of the relationship between the whole and its parts, plus the idea that to find a missing part, you subtract the known part from the whole, therefore calculating 564 – 327 to find the unknown.

It is this kind of thinking that we must get children to do.  Maths lessons should be planned with a sole priority – what will the children be thinking about?  In this first stage of a unit of work, children develop their fluency and then begin to reason about what is unknown in a numerical problem and how to figure out that unknown.

Deliberate teaching of language

The second stage of my model involves building on the number work by adding layers of mathematical language that enable children to talk like mathematicians and understand problems involving ambiguous language. We need to embrace ambiguity because it is in that grey area of language that we can really get children to think hard.  The image below is one I’ve seen many times in many schools and even had up in my classroom in a previous life:

What this kind of display tells children is that a word equates to an operation.  This is misleading at best and more likely disastrous for understanding.  Words like this only carry meaning in context, for example look at the phrase ‘more than’:

It does not mean that you have to add the numbers together.  It can mean that, but it could also mean finding the difference.  And what of the word difference?

The ambiguity of language must be deliberately taught and linked back to the underlying structures that children will have been working on.  Teachers model the thinking and ask: In the first  question, is the whole unknown or a part unknown?  What about the second sentence? Draw it…

Wider problem solving

Remembering Hattie and Yates’ assertion that novices see surface features of problems and experts see the same problems at a deeper level, consider this:

Anum and Jay have saved up their pocket money.  Altogether they have £35 and Anum has saved £18.50.  How much has Jay saved?

A novice would read that question and would say it’s a problem about pocket money – the surface feature.  An expert would look at the same problem and say that it’s one where the whole is known and so is one of the parts, but that the other part is unknown.

Once teachers know how experts think, it is perhaps a mistake to simply try and get novices to think like that.  Experts have a vast store of knowledge from which they draw on when analysing problems and so to get novices to eventually think like experts, we must first teach them the underlying patterns.

With this in mind, take a look at this common practice:

If we teach children how to subtract, then give them problems that only require subtraction, what are children really thinking about?  Not structure.  Not language.  And yet we can still mislead ourselves when children ‘get the right answers’ that they truly understand what they’re doing.  They may well do but we can’t be sure with tasks like this.  The tasks that we set show what we value.  Perhaps a better task is to ensure that children are thinking like mathematicians, sorting then solving problems based on their underlying structure:

Throughout the entire unit

Other considerations during a unit of work include the big ideas in maths.  Coined by Mike Askew, these are concepts that children develop throughout their time at school and are built on year on year:

• position on a number line
• estimation
• equivalence
• place value
• numerical reasoning
• the meaning of symbols
• classifying
• sequences

Opportunities should be created throughout a unit for children to think about content in these ways so that they can make connections between ways of thinking and different representations.

Finally, but by no means of least value, teachers must pay careful attention to success criteria. Before this though, objectives need to be sound. Shirley Clarke’s work on formative assessment is important here and there should be clarity between closed and open objectives. Closed objectives are absolute. They have either been achieved or they have not. Procedural success criteria are most appropriate here – steps to follow in order to be successful. Open objectives on the other hand are subjective in that they can be achieved to varying degrees of quality. A selection of tools is most appropriate for success criteria in this instance – strategies to choose from with the goal of efficiency, for example. Year 5 children could be given the calculation 5023 – 3786 and they should be able to, following great teaching, choose a subtraction strategy to solve that calculation in the most efficient way possible. All the procedural work, such as deliberately learning to count up, count back, round and adjust, or carry out column subtraction comes together and all those strategies form a toolkit from which children choose he best tool for the  job.

This model for teaching maths certainly covers National Curriculum aims but more importantly, it strives to get children to think and communicate like mathematicians.

Filed under Maths

## The teaching of fractions

There are certain prerequisites for children to develop a solid understanding of fractions.  First, they must understand, through work on additive reasoning, that a whole can be split into parts and that the sum of those parts is the whole.  There’s a short step into multiplicative reasoning from here – that a whole can be split into multiple, equal parts and that the whole is the product of the size of each part and the number of parts.  Once this is understood, children can begin to think about the whole being worth one and the parts being fractions of one.  The ideas that follow are broadly sequential in terms of conceptual development.

Children will need to manipulate various representations of fractions, for example making them with fraction tiles (as both bars and circles); taking strips of paper and ripping them in to equal parts; and drawing bars and circles, dividing them into equal parts.  It is worthwhile to get children to do lots of judging by eye and marking equal parts of a whole as well as using squared paper to do so accurately.

Of course, there is a lot of language to work on whilst manipulating these models of fractions.  Children need to be shown clearly the link between the total number of parts and the language (but not yet necessarily the written form) of the denominator: two parts – halves; three parts – thirds; four parts – quarters etc.

With a secure start in the basics of splitting a whole into equal parts, children can work on the idea that fractions always refer to something.  A third, for example, doesn’t stand alone.  It might be a third of an apple or a third of twelve sweets or a third of one whole.  Modelling these full sentences and getting children to speak in this way should solidify their understanding of proportion.  Through the sharing out of objects, even very young children can work on the concept of fractions of numbers – sharing six sweets between three children means that each child has the same number of sweets and that two sweets is one third of six sweets.

Once children are comfortable with the idea that an object or a set of objects or a number can be split into equal parts, and that each of those equal parts can be described as a fraction of something, that object or that set of objects or that number, they can go on to work at greater depth.  By comparing strips of paper or bar models that are the same length yet are split into different fractions, children can look at the relationship between the size of each part and the number of parts.  That is, the greater the number of equal parts, the smaller the size of each part.  Children should be expected to think about how ¼ is smaller than ½ because ¼ of one whole is one of four equal parts whereas ½ of one whole is only one of two equal parts.  Then, questions like this should be relatively straightforward:

The understanding that unit fractions with larger denominators are smaller than unit fractions with smaller denominators will contribute significantly to work in comparing fractions later on.

Children could begin to look at improper fractions and mixed numbers next.  Using ¼ fraction tiles, they could make one whole and then see what happens if you add another ¼.

This lends itself to counting in unit fractions but we should exercise caution.  Children may be able to chant ‘Three quarters, four quarters, five quarters…’ but early conversion to mixed numbers as well should help to secure their understanding of the relationship between them.  Manipulatives like fraction tiles and multi-link cubes are great for representing improper fractions because they can trigger accurate mathematical talk to describe the improper fraction (the total number of cubes as the numerator and how many cubes in each whole as the denominator).  The same can be done to describe the mixed number (the number of wholes, then what is left over as a fraction of a whole).

Returning to additive reasoning, children could generate complements to 1 whole and record them as addition and subtraction statements.

A slight change to the representation used can support children to work with complements where denominators are different:

Placing two bar models of equal length one on top of the other is great scaffold for comparing fractions.  When the denominators of the fractions are the same, the bars should not even be necessary but when they are different, the image can help to structure thinking.

When dealing with fractions with different denominators, the practice that children had earlier of judging by eye to split a whole into equal parts and marking the divisions themselves becomes crucial, otherwise, things like this could happen:

A standard fraction wall is all that is needed to begin work on equivalence and the first step is of course shading one fraction and looking up or down the fraction wall to find fractions of equal size.  When children are comfortable with that, they can begin to look at patterns in the abstract representations, particularly the link between times tables, numerators and denominators.

Using the language of simplifying or cancelling fractions without first talking more generally about the concept is a mistake.  If children are well versed in using a fraction wall to find equivalents to a given fraction, it is only a slight tweak to talk about finding the equivalent fraction that has the fewest total parts.  It would be tempting to talk about finding the equivalent fraction that is ‘closest to the top’ of the fraction wall but this would be a mistake too.  The language of simplifying or cancelling can be used to attach to the concept of finding the equivalent fraction with the fewest total parts to get children thinking conceptually soundly.

One further aspect of thinking of fractions is to consider them as numbers.  To do this, plotting fractions on a number line directly beneath the bar model is a good way of linking the two representations.

Representing fractions as a proportion of one, as a part of a quantity and as a position on a number line significantly supports children’s development of proportional reasoning and ensures that future tricky concepts such as calculating with fractions can be built on a secure foundation.

Filed under Maths

## How can a child catch up to learn times tables in one term?

Children should know all times tables by the end of year 4, but there are children that slip through the net, taking much longer to learn them.  There are also children that may seem to have learned times tables by the end of year 4, but forget and have to work into upper key stage 2 to relearn.

This post describes a plan to get children who are in year 3 and 4 and who are not on track to understand times tables by the end of year 4.  The plan is also for children in year 5 and 6 who still do not know their times tables.

A fact a day for a term

The basic structure of the plan is to work on one fact per day.  Working with commutative facts such as 3 x 4 and 4 x 3 together, and taking into account that familiarity with tasks should accelerate the work the longer it goes, a term is a sensible time frame to work in.  This will be systematic, working from x10 to x5, then x2, x4 and x8, then x3, x6 and x9, finishing with x7, x11 and x12.  This is to enable links to be made between times tables.  Within each times tables, we’ll work in increasing order of times tables (i.e., 10 x 1, 10 x 2, 10 x 3 etc.).  Of course, different children will have different starting points, not all starting with 10 x 1.  As days pass, children will consolidate their understanding of a times tables through repetition, multiple representations, counting and low stakes testing.

Multiple representations

For times tables to stick and to be useful in other areas of maths, they need to be rooted in secure understanding.  To allow this to happen, each fact will be represented in different ways, in the first instance by the teacher but increasingly by the child.  The first representation is Numicon, using the example of 4 x 5:

Using this we can explain that 4 x 5 means 5 lots of 4 and that by counting in multiples, we can find out that 4 x 5 = 20.  Children will have done this for 4 x 1, 4 x 2, 4 x 3 and 4 x 4 in the preceding days so they should be able to count in 4s.  However, they may need to do some skip counting, where they whisper or say in their head each number except for the last on each Numicon piece (1, 2, 3, 4; 5, 6, 7, 8; 9, 10, 11, 12…).  The Numicon also helps to lead into other representations:

Repeated addition: 4 + 4 + 4 + 4 + 4 = 20

Bar model:

Number line:

All the while, the child is practising counting in 4s, and thinking about how 4 x 5 = 20.

Commutativity

One more representation can lead the child into working on the related commutative fact.  An array gives a little further practice seeing how 4 x 5 =20:

Rotating the array shows how 5 x 4 has the same product:

This can lead into counting in 5s to get to 20 and showing that 5 + 5 + 5 + 5 = 20.  Then, repeating the representations of Numicon, a bar model and a number line will help to internalise the commutative fact.

Low stakes testing

Having worked on this new fact (and its commutative relative), the child can then work on remembering facts that have been previously worked on in days gone by.  Practising recalling times tables is of course a great way of ensuring that they come to mind immediately when needed.  Quick, effortless recall means that little cognitive effort is required to summon the knowledge, thereby keeping as much working memory as possible freed up to solve a problem that needs the times table fact in the first place.

There are two ways of working on quick recall of times tables.  The first is if the child has a reliably secure understanding of multiplication.  In this case, simple testing such as asking ‘What is 3 x 5?’ or the use of individual flash cards will be fine.  However, if a child is still not quite there with conceptual understanding, testing by using objects or images can help to get them to think mathematically instead of guessing.  The teacher shows any of the pictorial representations already described to prompt thinking about the number of groups, the size of each group and ultimately quick recall of the whole.

1 Comment

Filed under Maths, Uncategorized

## Multiplicative reasoning

If a child understands additive reasoning and the relationship between the whole and its parts, it is a fairly straightforward conceptual step to understand multiplicative reasoning.  Multiplicative reasoning should be modelled as repeated addition in the first instance.  Adding multiple equal parts (for example 5) might look like this:

5 + 5 + 5+ 5 is equal to twenty.  Children need to understand that multiplication allows for efficient repeated addition.  You  have your thing to be multiplied (5) and the multiplier (4): 5 +5 + 5 + 5 = 5 x 4.  Creating arrays and deliberately connecting repeated addition with multiplication makes for sound understanding.

How children work out the whole should not be taken for granted.  At first, children might count each item in the array.  Counting in multiples can be achieved by first skip counting.  Children might whisper the numbers while counting except for the last in each row, which is said out loud.  Then replace the whispering with counting in their heads and then simply saying the multiples.  Over time, given sufficient practice, children will internalise these times tables.

Commutativity is important here – the array used above shows 5 x 4 but rotated it shows 4 x 5.  Times tables taught systematically and with such conceptual support should be straightforward for children to learn comfortably before the end of year 4, especially when we consider it like this:

Of course, children need time to practise well and multiple representations help children to make connections.  Graham Fletcher’s blog post describes the use of pictorial representations on flash cards – an approach that is a great form of low stakes testing to support the learning of times tables.

This image supports the understanding of having a ‘thing to be multiplied’, a multiplier and a whole.  With practice, children will be able to subitise from glancing at the flash card, becoming fluent and accurate with times tables recall.

Some children will grasp all this quickly and can work at a greater depth while children that need more practice with the basics get it.  Still using the array, children can easily begin to think about distributivity simply by splitting the array into parts:

The part above the line is 5 x 2 and the part below the line is 5 x 2:

5 x 2 + 5 x 2 = 5 x 4.

There is lots of scope for systematic thinking about equivalence with a task like this.

Arrays are perhaps not the most efficient of representation so a progression is to get children to be able to represent multiplication in bar models.  First though, Numicon to work on the language of size of each part, number of equal parts and the whole:

Numicon is a great manipulative to represent multiple parts because of its clarity of the ‘size of each part’.  Multi-link cubes could work too, but children would need to organise the parts into different colours to differentiate between them:

Building worded statements using a manipulative will ensure children practise the language needed to internalise the concept of multiplicative reasoning.  Dropping in some of the  inverse relationship between multiplication and division could be useful here too.  Doing it systematically can also help keep times tables knowledge conceptual and not shallow:

Commutativity could be brought in again – showing that 3 groups of 4 is the same as 4 groups of 3 using manipulatives arranged with intent.  Alongside this, comparing the similarities and differences with the worded statements will get children to think with clarity about equivalence between two multiplicative expressions.

Bar models are a versatile representation that can be used to solve a wide range of problems later on, so getting children to sketch out multiplication and division statements using bars enables them to practice a versatile skill.  We should expect great accuracy in their drawings – they should be representing equal parts.  If children also represent the same expression on a number line beneath the bar model, we can encourage links between representations and lay the foundations for trickier calculations and problem solving as they progress through school.

Update: The NCETM recently published this account of teaching the six times table, with some great ideas for depth.

1 Comment

Filed under Maths

Could Reception and Year 1 children solve this problem?

4 + 3 = 2 + □

Of course they could.  Here’s how.  First children will need to work on their understanding of 7.  Using a manipulative for 1:1 correspondence such as multi-link cubes, we can show how the whole of 7 can be made up of two parts (in the first instance, 1 and 6):

It is important to model the language that will help children think clearly when manipulating the cubes: ‘One add six is equal to seven.  The parts are one and six and the whole is seven.’  It is equally important to talk about the cubes saying the whole first: ‘ Seven is equal to one add six.’  This will help to prevent the misconception developing that the equals sign means ‘the answer is next’.  Then show them how to systematically make seven with other sized parts, talking about the parts and the whole in the same way:

Children should also use the cubes to write calculations.  A little modelling of turning the language of ‘Three add four is equal to seven’ into 3 + 4 = 7, followed by plenty of practice, will be exactly what is needed.

Lots of quality talking, as well as using pictorial representations, will develop children’s fluency with number facts.  Showing different representations, for example Numicon, could strengthen their conceptual understanding:

Some children will grasp this idea quickly, and some will need more practice to internalise the number facts and recall them more fluently.  Those quick graspers can be challenged to think more deeply about the number facts that they are working with.  We can start by returning to the multi-link cubes and looking at two facts:

Here, we can model the talk required to think more deeply: ‘Three add four is equal to five add two.’  Children could repeat that task with different facts to 7 before we show them how to write that as 3 + 4 = 5 + 2.  When children have practised this and can do it reliably with manipulatives, they could draw a bar model of what is happening:

A further challenge is to present cubes where there is an unknown:

We could model how to talk about this as: ‘One add six is equal to three add something.’  To model how to work out what ‘something’ is equal to, we simply fill the gap with cubes to make the second row equal to seven, then counting the cubes to figure out what ‘something’ is equal to.  When children have practised and are becoming more fluent, the cubes could be replaced with bars, at first presented in that way but moving on to children drawing it themselves:

All the while, children could be shown how this looks written down: 1 + 6 = 3 = □.  When they have seen the abstract alongside the pictorial and the concrete, we can try starting with the abstract and asking children to represent the problem with cubes or by drawing bars.

The sequence described, over time, should be enough of a scaffold for the vast majority of children to end up being able to solve such problems and in doing so, develop a deep understanding of early number.

Filed under CPD, Maths

## Enumerating possibilities of combinations of two variables

With Year 6 children expected to work on the objective ‘enumerate possibilities of combinations of two variables’, we should be clear on the difference between the underlying concept and the algebraic representation of it.

2g + w = 10

For questions such as this, children should first have a secure understanding of the part, part, whole model.  We can show that 2 lots of something add one lot of something else is equal to 10 by using a concrete manipulative such as Numicon.  First, children represent the whole, in this case 10. Then they can speculate on the two equal parts (2g), trying out g=1 before finding the Numicon piece that fills the gap and therefore is equal to w:

Having found one solution, they can continue to work systematically to find alternative solutions.  Trying g = 2 is logical:

Lining up solutions beneath the whole reinforces the idea that the expressions are equivalent.  Children can continue to work systematically:

This also provides a scaffold for questions of greater depth, such as ‘What is the greatest number that g can represent?  Explain…’
Subtraction?  Not a problem, although in this case, children must know that for subtraction, you always do so from the whole.

10 = 3g – w

In this question, the whole is 3g and the parts are 10 and w:

What is not clear from this model is the trial and error that went into it.  Children may well try 3 ones and quickly realise that it is already less than 10, so subtracting from it will not give a valid solution.  There is lots of scope here for discussion about the smallest number that g could represent.

The use of Numicon leads nicely into children representing problems as bar models.  Here are the two examples used so far:

Filed under Curriculum, Maths

The question was on the screen:

One year 6 child said: ‘The empty box is in the middle so you do the inverse.  You have to add the numbers together’.

This got me thinking about how children build on their early concepts of number to be able deal with problems like this, which I’ll call ‘empty box problems’.

The underlying pattern of additive reasoning is the relationships between the parts and the whole.   Getting children to think and talk about the whole and parts using concrete manipulatives early on should lay the foundations for them to internalise this underlying pattern.  Every time children think and talk about number bonds, they can be practising identifying the whole, breaking it into parts and then recombining to make the whole once more.

Alongside talking about the whole and parts, children should begin to generate worded statements whilst manipulating cubes or Numicon, for example.  At this point it is important to experiment with rearranging the words in the statement.  They should get to know that ‘four add two is equal to six’ and ‘six is equal to four add two’ are statements that are saying the same thing.  Some discussion around what is the same and what is different about these two statements would be worthwhile.

When children are then shown how this looks abstractly with numerals and the equals sign, this would hopefully go some way towards avoiding the misconception that the equals sign means that ‘the answer is next’.

In the examples used so far, the whole and each of the parts have been ‘known’.  Using the same manipulatives and language patterns, children can be introduced to unknowns.  It seems sensible to begin with giving children the parts and using the word ‘something’ to show that the whole is unknown, i.e., four add two is equal to something.  Some modelling alongside a clear explanation followed by plenty of practice should see children get used to the language patterns needed to think about the concept with clarity.  The next step is to show children the whole and one of the parts, using the word ‘something’ to replace the unknown part.  All of this talk and manipulation of objects is intended to support children to develop a concept of additive reasoning where they do not have the misconception that ‘inverse’ means ‘do the opposite’.

More sophisticated additive reasoning is the understanding of the inverse relationship between addition and subtraction.  Children need to fully understand that two or more parts can be equal to the whole.  From this, they need to internalise the underlying patterns: that Part + Part = Whole and that Whole – Part = Part.  From this, they should be able to work out the full range of calculations that represent one bar model.  Again, it is important to vary the placement of the = sign.

One more way to get children to think about the whole and the parts is to use bar models for calculation practice rather than simply writing a calculation for children to work out.  When done like this, children have to decide what calculation to do to work out the unknown.  Children often exhibit misconceptions such as ‘when you subtract, the biggest number goes first’.  These can be addressed using the underlying patterns; adding parts together makes the whole and, when you subtract, you always subtract from the whole.  When unknowns are introduced, they can be substituted into these basic patterns:

Part + Something = Whole           Part + □ = Whole              35 + □ = 72

Something + Part = Whole           □ + Part = Whole              □ + 35 = 72

Whole – Something = Part           Whole – □ = Part               72 – □ = 35

Something – Part = Part                □ – Part = Part                   □ – 35 = 37

Knowing these patterns will help children to able to analyse problem types in order to decide on the calculation needed.  An additive reasoning bar model with one unknown generates both an addition statement and a subtraction statement.  Showing children empty box problems pictorially, they can talk through the calculations that can be read from the bar model, using the word ‘something’ to represent the unknown.  The next step is to show children abstract empty box problems and get them to map it onto a blank bar model.  They should be drawing on their knowledge that the whole is equal to the sum of the parts and that when you subtract, you always start with the whole.  Eventually, the hope is that the language alone should suffice to work out how to solve empty box problems, with children no longer needing the bars.

Which brings us back to that year 6 child.  Of course, children will develop misconceptions as they make sense of what is shown and explained to them.  By expecting them to think and talk about additive reasoning in the ways described above, it should go some way to building sound conceptual understanding.