Pages

Monday, April 14, 2008

Fractions: An Example

We're up to 5 in our series of posts fleshing out the material written by Hung Hsi Wu in Critical Concepts for Understanding Fractions. See also Part I ,Part II, Part III, and Part IV.


Here, we go to a concrete example of all of the elements we've discussed: definitions of fractions, use of the number line, equivalent fractions, and some simple operations on fractions--comparing fractions and addition of fractions.

We consider Decimal Fractions. Decimal Fractions are a particular kind of fractions: fractions whose denominator is a power of 10: 10, 100, 1000, or any number represented by 10^n, where (for simplicity) n is a non negative integer.

here are some examples: 1489/100, 24/100000, 58900/10000

Decimal fractions have another convenient representation: as numbers that can be abbreviated into decimals: 14.89, 0.00024, 5.8900

Specifically, the above are finite decimals, though they are usually referred to simply as decimals. The number of digits to the right of the decimal point tells us the number of zeros in the denominator: 2 in 1489/100, 5 in 24/100000, 4 in 58900/10000. In this form, the convention is that zeros are added to the left of numbers as necessary to indicate the number of zeros in the denominator: in the case of 24/100000, we added three zeros to the left of 24: .00024.

The example 5.8900 shows that our decimal convention is not enough for explaining that we can remove the trailing zeros to the right of our number and decimal point. to show that, we need to invoke equivalent fractions:

for all whole numbers k, m, and n, where n and k are non-zero,
m/n = km/kn.


In this case, we can show that 5.8900 is the same as 5.89 by going back to the fraction form:

5.8900 = 58900/10000 = (589 * 100)/(100 * 100) = 589/100 = 5.89


This invocation of equivalent fractions works for in general, so any number of trailing zeros on the right can be shown to be equivalent to their absences:
12.700000 = 12.7 = 12.70 = .... etc.

After equivalent fractions, we moved on to comparing of fractions. Here, we show how to compare decimals.

Example: given 0.0082 and 0.013, which is bigger? We can compare these decimals easily by first converting them to fractions, giving them common denominators, and then comparing the numerators.

Converting to fractions, we see that we are comparing 82/10000 and 13/1000. To compare fractions, we then invoke again the rule that any two fractions may be represented by the same denominator:

m/n = ml/nl and k/l = nk/nl.


If m = 13 and n = 1000, then l = 10, and we have 13/1000 = 130/10000.Now we can compare 82/10000 and 130/10000 by comparing numerators. 130 is larger, so .013 > .0082.


Now, we move onto an application of fraction addition: to understand the algorithm for adding decimal fractions.

Example: 4.0451 + 7.28

Looking at our example and recalling what we know about fractions, the decimal fractions are just short hand for these numbers as fractions: 4.0451 = 40451/10000, and 7.28 = 728/100. We can then add them as we add any fractions, by giving them the same denominator. In this case, we call m = 728 and n = 100. If l = 100, then ml = 72800 and nl = 10000. Now, adding the whole numbers is the same as adding the numerator: we add

40451/10000 + 72800/10000 = (40451 + 72800)/10000 = (113251)/10000.

This leads us to an algorithm for adding decimals is:
1. line up the numbers by their decimal point
2. add the numbers just as you would "normal" whole numbers
3. put the decimal point in the number, based on where it "lined up" with the decimals in the additions.

When we line up the numbers by their decimal points, we are essentially rewriting the two decimals so they have the same number of digits to the right of the decimal point. This is the same as giving both of the decimal fractions the same denominator.

Then, we add the numbers, just as we would add the numerators.

Then, we determine the appropriate location of the decimal point just as we do by convention: we count the number of zeros in the denominator, and place the decimal point so that the number of digits to the right of the point matches the number of zeros in the denominator: 11.3251.

The point here is to stress that decimal fractions are really just fractions, and that we know how to perform elementary operations on fractions just as we do on whole numbers. Everything we understood about how to manipulate whole numbers on the number line mapped to fractions on the number line, and decimals are just a convenient notation for a certain type of fraction. The point is to underline how similar the processes are.

We'll get into more complications, like multiplication of decimals, later, after we advance a bit more with fractions themselves. Onto understanding fractions as division, then to multiplication and division operations!

7 comments:

  1. In the early grades decimal numbers are handled as decimal fractions: any number which can be written as a fraction with a power of ten for denominator. At some point the notion of infinite decimal expansion comes in. In fact, I think there were KTM questions posts at one point about converting between rational numbers (numerator, denominator) and their decimal representation. I know what that means, because I have taken calculus and understand about convergent series. But how is this done in high school algebra without the full machinery of convergence? Does anyone know of a good source for teaching this topic?

    ReplyDelete
  2. I'm not sure I undesrtand your question.

    You convert from a rational number represented as a fraction to a decimal by division. That can't be what you're asking. So you are asking how to convert from a (non repeating) decimal back into a fraction? huh? .382 is 382/1000.

    Are you asking about repeating decimals? i.e. how to go from .83333... to a fraction?

    .833333.... = x
    10*x = 8.33333....
    10*x - x = 8.3333 ... - .833333 = 7.5
    9x = 7.5
    18 x = 15
    x = 15/18 = 5/6

    Are you asking how to justify to an 8th grader that you can subtract off the infinite part? Or are you asking something else? I don't see how convergence comes up at all. I see that .3333333333333...
    is 1/3. and that once you've learned that rote, it shouldn't be a problem, so you can subtract the 1/3rds out of your problem jsut as you would any other fraction. What am I missing?

    ReplyDelete
  3. There are two kinds of rational numbers, those with a finite decimal expansion, like 1/4 (or anything with only 2s and 5s as factors in the denominator) and others like 1/3 which require an infinite (although repeating) decimal expansion.

    In other words, 0.333 has a simple definition. It means 333/1000. But 0.333... does not have a simple definition. To make full sense of anything infinite, you have to discuss the convergence of series.

    For eighth graders, there is going to be some compromise. How in fact is this done at that age? Does anyone know of a good textbook that does a good job of presenting this material?

    I like Wu's approach to fractions: be clear and thorough, use reasoning to explain why arithmetic works. In that sense it is easy to say what 1/3 is. But decimal representation are more sophisticated. Is there a way to teach decimals that handles the infinite aspect in a way that is consistent with Wu's development of fractions?

    ReplyDelete
  4. ---To make full sense of anything infinite, you have to discuss the convergence of series.


    No, I disagree. Divergence is a perfectly reasonable concept. Some things don't converge.

    In any case, repeating decimals are just fractions--and you know how to handle the fractions, so you learn the method (that I listed above), and then think about what it means as fractions.

    Wu's development is consistent with this, because it says "this decimal expansion stuff is just a representation--it's just a shorthand for notation."

    But I don't know of how high school texts treat the above problem and solution. I'd be interested in seeing what explanation they give for it.

    ReplyDelete
  5. Okay, so I guess I understand your point is that you don't know how to teach that 1/3 is .3333 repeating without showing that
    .3^i only reaches 1/3 as i gets to infinity.

    but okay--12th graders can handle that, and probably 11th too. I don't know how that's taught, but the visual representation would seem reasonable if they haven't yet learned series (which they do before they can do calculus)

    but for 5th-8th graders, who cares? There aren't "two kinds" of rational numbers. Rational numbers are defined as a/b where b is non zero. Decimal representation is merely a convenience. Fractions like 1/3 don't have a finite representation because they don't--and that's okay. we show that it repeats to show that we could go on forver, not that "if we reach forever, we're done."

    1/3 isn't 3/10 because 1/3 means to divide in threes. you manage to divide up 10 items into threes as : 3 into one pile, 3 in the next, and 3 in the next with one left over. same for 100: 3 sets of 33/100s leaves 1 left over. you don't really need to be showing that you're getting closer to 1/3, just that yo uall you need to point out is the right representation isn't in a real decimal, it's 1/3. with that, you could say we can go on forever trying to divide that last element into thirds. You don't have to show that if magically you ever got to that infinitely last element, you'd gotten to a third.

    ReplyDelete
  6. sufficient onto the day
    is the rigor thereof ...
    and while, yes, a *formal* treatment
    of "infinite decimals" requires
    a definition of convergence,
    on the other hand, it's probably
    a *bad* idea to work at this level
    of formality before, say, algebra.
    the founders of calculus did fine
    without defining convergence;
    elementary school children
    the world over have been comfortable
    for generations with expansions like
    1/3 = .333333...
    and shouldn't be urged to become
    *uncomfortable* with 'em because
    some egoface wants to show off
    a bunch of college-level analysis.

    this is *not* to say that we should
    let textbooks continue to get away
    with simply *asserting*, for example,
    that "ratios of integers"
    and "repeating decimals"
    name the same set -- this is easily
    proved (allison has sketched
    the hard step [decimal to ratio];
    the easy step [ratio to decimal]
    is of course just The Divison Algorithm.

    oops, gotta go. late for tutoring.

    ReplyDelete
  7. At least in Saxon II, there doesn't seem to be an explanation of why. It tells you that it is the case that all repeating decimals can be represented by fractions and then, as you say they illustrate how you can add and subtract the infinite part using a particular repeating decimal. At least in the section I am looking at there is no proof, no generalization, just a "how to"

    In Jacob's Algebra, it is treated in one problem set like this, "Most rational numbers do no come out "even" if changed to decimal form, but "keep going," For exxample, 1/3 = 0.3333..., in which the three dots indicate that the threes continue without end. It is obvious from this pattern that the 100th digit after the decimal point will also be a 3."

    Gotta love that handwaving!

    Foerster's Alegebra, the one rated the highest by mathematicallycorrect.com has no explanation of repeating decimals that I could find. The only explanation was how to round decimals on a calculator.

    Singapore's NEM is for 12/13 year olds and teaches this by giving a single example of a long division problem with a repeating decimal and then tells the student, "Notice that the division is not exact. There is always a remainder. After some steps we come to the same remainder. It is clear that the set of digits '45' in the quotient repeats endlessly. Thus we write the quotient blah blah blah.


    Art of Problem Solving has an entire chapter on decimal expansion in their number theory book and has one sidebar which might have what you seem to be looking for, "Another method for converting rpeating decimals to fractions involves summing an infinite geometeric series. We rework problem bla blah blah this way.."

    The old 1960s book we are currently using with our son begins a two page discussion of repeating decimals like this, "Suppose an obgject starts at the orgin of the number line and moves to the right 10 units during the first second, 1 unit during the second second, one tenth of a unit during the third second, one hundredth of a unit during the fourth second, and so on. It is evident that the distance d which the object travels increases with time. Does this mean that the object will eventually pass the point whose coordinate is 12? Let us investigate the situation...

    If you want to see how this ties in to repeating decimals you can see both pages here:

    http://i231.photobucket.com/albums/ee27/Carrie3d/RepeatingDecimals.jpg?t=1208470337

    ReplyDelete