Data Analysis

If there’s one thing we can say about modern society, it’s that we
love our data. If you spend just a few minutes on the Internet, you
can find out roughly how many people there are in America; how
much, on average, they earn; how long they’re expected to live;
what percentage of them marry; what percentage have kids; how
many kids they have; how much those kids weigh; and so on and
so forth until every aspect of daily life is reduced to numerical
form. Perhaps expressing everything numerically makes us feel in
control of our fate, or perhaps it helps us cope with the
ambiguities of our complex technological lives. Whatever the
reason, we love our data, we surround ourselves with it, and we
have invented all sorts of statistical mechanisms to express what
all this information means.
The GRE test makers have taken our numerical fetish to heart. In
this final Math 101 section, you’ll find more than you ever wanted
to know about statistical concepts such as mean, median, mode,
probability, and every other data analysis topic tested on the GRE.
Mean
On the GRE, mean and arithmetic mean both represent the
concept that you may recognize by its more common name,
average. No matter what we call it, the calculation is the same:
Add up all the terms and divide by the number of terms. You’ve no
doubt seen this in school: If you get scores of 90, 95, and 100 on
three tests, then 95 is the average of the three test scores. In this
basic example you can probably see at a glance that 95 is the
average, but technically you can calculate it by taking the sum of
the scores (90 + 95 + 100) and dividing it by the number of scores
(3). The formula, in general terms, is:

Let’s try one out:

What’s the arithmetic mean of 3, –5, 7, and 0?
Solve by using the formula:
Some mean problems may be straightforward like the one above,
but the more complicated ones may give you two values and ask
you to solve for a third. For example, the test makers might give
you the mean and the number of terms and ask you to solve for
the sum of the terms. Your job will still be to plug the known
values into the formula and solve from there. Here’s an example:
The average height of five people is 54 inches. One of the people
leaves the group, and the average height of those remaining is 52
inches. How tall is the person who left?
In the first sentence, we’re given the number of people, five, and
their average height, 54. We can use the mean formula to
calculate the sum of the heights of these five people:

In the second sentence, we’re told that the average height of the
remaining people is 52. Since one person left the group, four
people remain. Plugging 4 and 52 into the mean formula gives:

The difference between the sum of the heights of the original five
people and the sum of the heights of the remaining four must be
equal to the height of the person who left. Subtract the second
sum from the first to get the height of the person who left: 270 –
208 = 62 inches, the final answer.
Median
The median of a group of numbers is the middle term when the
numbers are written in either ascending or descending order.
That means that before you can calculate a median, you must first
rewrite the terms of the group in ascending or descending order.
For example, to calculate the median of 0.3, 7, 0, 9, and 10, you
can’t choose 0 simply because it appears in the middle. You must
first write the numbers in order: 0, 0.3, 7, 9, 10. Since 7 appears in
the middle of this ordered list, 7 is the median.
If two numbers appear in the middle, which will happen whenever
the total number of terms is even, take the mean of the two
middle numbers to determine the median. For example, the
median of 1, 2, 4, and 8 is 3, since 3 is the mean of 2 and 4.
One wrinkle you may come across in a median problem is
a description of a list of consecutive numbers, instead of a list of
the actual numbers, as in this example:
What is the median of all the integers between 210 and 260,
inclusive?
You certainly don’t want to write out all the numbers from 210 to
260, and then try to find the one in the middle. It’s better to use
the following formula:

In our example, the median would be:

Remember, inclusive means “including,” which is why we used

210 and 260. Had the question said exclusive, we would have used
211 and 259, divided by 2, and also gotten a median of 235.
Mode
The mode of a group of numbers is the number that occurs most
frequently. If multiple numbers are tied for first place in the race
to occur the most, then the group will have more than one mode.
For example, the modes of the set of numbers {6, 6, 1, 3, 4, 1} are
6 and 1, since both 6 and 1 occur twice, and all the other numbers
occur only once.
Range
The range of a group of numbers is the difference between the
largest term and the smallest term. For example, the range of 10,
–25, 3, 2, and 4 is 10 – (–25) = 35. One would need to travel 35
units on a number line to get from the smallest value, –25, to the
largest value, 10.
Standard Deviation
Standard deviation is one of the most difficult statistical concepts,
but thankfully you’ll only need a very general understanding of it
for the GRE. The test makers won’t ask you to actually calculate
standard deviation, as the formula for doing so is pretty difficult.
You will, however, be expected to know that standard deviation is
a measure of how spread out a group of numbers is. The more
spread out a group of numbers, the larger its standard deviation.
Let’s look at an example:
Which of the following groups of numbers has the greater
standard deviation?
Group A: 11, 12, 13, 14, 15
Group B: 50, 51, 51, 52, 53
Even though the numbers in Group B are larger than those in
Group A, they’re closer together thanks to the double occurrence
of number 51. No such overlapping occurs in Group A. Group A
exhibits a slightly greater spread and therefore has the greater
standard deviation.
The 34-14-2 Rule
Standard deviation also lets you know how likely it is that a value
will differ from the mean by a certain amount. In general, the
farther a value is from the mean, the less likely it is to occur. The
following graph, called a Normal Distribution, shows this in more
detail:

This graph is the basis for the 34-14-2 Rule:

34%, 14%, and 2% represent the likelihood that a value will fall
into each given region. For example, there is a 34% chance that a
value will be between the mean and one standard deviation to the
right or left of the mean. Similarly, there is a 14% chance that a
value will be between one and two standard deviations to the right
or left of the mean.
Here’s how to use the numbers. Say that the mean for some group
of values is 10, and the standard deviation is 2. One standard
deviation to the right of the mean will therefore be 10 + 2 = 12.
The Normal Distribution graph states that there is a 34% chance
that a value from the group will fall between 10 (the mean) and 12
(one standard deviation up from the mean). Based on this kind of
analysis, you may be asked something like this:
What is the likelihood that a value within a group of values with a
mean of 10 and a standard deviation of 2 equals 5?
If the mean is 10, then one standard deviation below the mean is
10 – 2 = 8, which creates a 34% chance that a value from the
group falls between 10 and 8. A second standard deviation to the
left would be 8 – 2 = 6, meaning that there’s a 14% chance that a
particular value would fall between 8 and 6. More than two
standard deviations to the left of the mean would be all values
below 6. The graph tells us that these values have a 2% likelihood.
The number 5 falls into this group, so 2% would be the answer.
Frequency Distribution
Say, for example, that fifteen college graduates were asked how
many different jobs they had in their first five years following
college, and the responses came back like so:
2, 4, 2, 1, 0, 1, 3, 2, 2, 5, 3, 4, 1, 2, 1
Not very pleasing to the eye, is it? One way to organize this
information is to express it in the form of a frequency
distribution: a chart that shows at a glance all of the answers
given and the number of people who gave each answer. Frequency
distributions typically designate x as the values (in this case, the
answers given by those surveyed) and f  as the frequency of each value. In the
example above, that would look like this:
x f
0 1

1 4

2 5

3 2

4 2

5 1

Total 15

We can see from the chart that one person surveyed had no jobs
in the first five years after college (slacker!), four people had one
job, five people had two jobs, and so on. So what’s so great about
this? Well, the best thing about it is that it allows us to quickly
determine many of the other statistical features we’ve been
discussing so far. For example, eyeing only the left-hand column
tells us that the range of responses is 5 – 0 = 5. A quick scan of the
right-hand column indicates that 5 is the largest frequency
corresponding to any one answer, and it corresponds to the
answer 2, which therefore qualifies 2 as the mode. The chart
already lists the responses in ascending order, so the median will
be the eighth value from the beginning—eighth because with
fifteen values total, the eighth value is right in the middle with
seven values below it and seven above it. The first value is 0, the
next four values are 1, bringing us to the fifth value, and the next
five values are 2, bringing us to the tenth value. The eighth value
is therefore a 2, so 2 is the median of this group of values.
You may also be asked to calculate the mean from a frequency
distribution. Recall the formula for mean:
Using the frequency distribution, we can quickly calculate the sum
of the terms by finding the sum of each term multiplied by its
frequency. The number of terms will be either given in the table
(such as “total = 15” in our chart), or you can just add up the
frequency numbers in the right-hand column to calculate the
number of terms. In this case, the mean is:

If you learn the basics of frequency distributions, it should be a

welcome sight if one of these appears on your test.
Probability
Probability is the measure of how often something is expected to
occur, expressed as a fraction or decimal between 0 and 1. A
probability of 0 means there’s no chance that the event under
consideration will take place. A probability of 1 means
itdefinitely will happen. Most probabilities tested on the GRE fall
somewhere in between. We’ll use the common scenario of
selecting colored marbles from a bag to illustrate the various
kinds of probability questions you might see on your test.
Single Trials
The most basic kind of probability question involves a single
selection from a given group of elements. Here’s an example:
In a bag containing 12 red, 13 white, and 15 black marbles, what
is the probability of selecting a red marble on a single draw?
To tackle probability problems, use the following formula:

The number of favorable outcomes is math lingo for the number

of ways you can get what the problem is asking you to get.
Here, red marble is the favorable outcome, so the numerator of
the fraction is 12, the number of red marbles in the bag. The total
number of possible outcomes is the total number of possibilities,
or, in our problem, the total number of marbles. Make sure to
include all of the marbles, including those already counted as
favorable outcomes. The total number of marbles is 12 + 13 + 15 =
40. Plugging 12 and 40 into the formula gives:

Simplifying this gives  , the final answer.

Independent Events
If one event does not influence the occurrence or nonoccurrence
of another event, the two events are independent. To find the
probability of two independent events occurring, simply multiply
their individual probabilities. For example, if there’s a 1 in 4
chance that Mary will be selected for a committee, and a 1 in 3
chance that Bill will be kicked out of college, and the events are
independent (that is, Mary isn’t angling to join the committee
with the purpose in mind of booting Bill), then there’s a 
chance that both Mary will be selected for the committee and Bill
will be given his college walking papers.
We mention independent events at this point because this concept
affects our next topic, multiple trials.
Multiple Trials
Frequently, probability questions on the GRE won’t be limited to
a single draw, or trial, but will instead involve repeated draws.
When a question involves drawing multiple times from the same
group of entities, you need to distinguish between draws with
replacement and draws without replacement. Let’s illustrate the
difference using our marble example:
1. 1 You select a marble, note its color, and put it back in the
bag. You then select a marble again. This is called drawing
with replacement.
2. You select a marble and put it aside. Then you draw another
marble from those remaining. This is called drawing
without replacement.

The GRE will always make it clear which method is being used
either by including the actual phrase with replacement or without
replacement or by explicitly describing the method of selection in
a way that makes it obvious which mechanism is in play. Let’s
look at an example of each type.
DRAWING WITH REPLACEMENT
Try your hand at this one:
A bag contains 12 red, 13 white, and 15 black marbles. What is
the probability of selecting two black marbles in a row if the
selection is made with replacement?
The number of black marbles, or favorable outcomes, is 15. The
total number of marbles is 40. First, use the probability formula
to find the probability of selecting a black marble on the first
draw:

Since this problem involves drawing with replacement, we’ll need

to put the black marble selected on the first draw back into the
bag before selecting again. So the bag will still contain 15 black
marbles out of 40 total for the second draw. The probability of
drawing a black marble on the second draw is thus the same  .
Now, even though the marbles are coming from the same bag,
these two events—a black marble on the first draw and a black
marble on the second—are independent; that is, what happens on
one draw doesn’t affect what happens on the other. To get the
probability of two black marbles in a row, we can therefore
multiply the individual probabilities:

DRAWING WITHOUT REPLACEMENT

Now let’s see what happens when we don’t put the first marble
back into the bag after selecting it:
A bag contains 12 red, 13 white, and 15 black marbles. What is
the probability of selecting two black marbles in a row if the
selection is made without replacement?
The probability of the first marble being black is  , just as before.
For the second draw, however, only 14 black marbles remain out
of 39 total. (Remember, we took a black marble out of the bag
and did not put it back.) This means that the probability of the
second marble being black is  . By assuming the first draw was
favorable (a black marble selected), we adjusted the figures for
our second probability. Since these figures are already adjusted to
account for the first favorable outcome, the second drawing is
independent from the first, so we can still multiply the individual
probabilities to get the chances of selecting two black marbles in a
row:

We can cancel the 3 and 39 before multiplying and also cancel a

factor of 2 from the 8 and 14 to make our lives easier:

This can’t be reduced any further, so it’s the final answer.

The Probability of Something NOT Happening
You’re probably familiar with the phrase not happening, as in
when you ask your boss for three weeks off in the summer and he
tells you “that is so not happening.” But do you know that there’s
an actual math formula for “not happening”? You probably do,
without even realizing it. For example, if you’re told that the
chance of snow tomorrow is 25%, it’s likely you recognize without
much thought that the chance that it will not snow is 75%. Here’s
the formula you used, whether you were aware of it or not:
the probability of an event NOT happening = 1 – the probability of
that event
This formula can turn very hard probability questions into easier
ones. Consider this next one:
A bag contains 12 red, 13 white, and 15 black marbles. What is
the probability of selecting at least one red or one white marble in
two draws if the selection is made with replacement?
This is harder than the previous problems, because it’s not
altogether clear what must happen on any individual draw for a
favorable outcome. For instance, the first draw might be black,
and you still could have a favorable outcome if the second draw is
red or white. Similarly, the second draw could be black, and you’d
still have a favorable outcome if the first draw is red or white. And
of course, a first and a second draw of red or white would also
count as a favorable outcome. So how do we deal with this
ambiguity?
Simple: Use the formula for “NOT happening.” It’s far easier in
this case to calculate the probability of not getting at least one red
or white marble in two draws because this is actually the same
thing as drawing two black marbles, with replacement. We
already calculated this earlier as  . The probability of drawing at
least one red or white in two draws is 1 minus the probability of
that NOT happening, which is simply the probability of drawing
two black marbles. The answer is therefore:

Still no piece of cake, but doable.

“Or” Questions
A difficult question may ask you for the probability of event
A or event B occurring, which is different from the probability of
A and B occurring. If you see one of these, use the formula:
probability of A or B = probability of A + probability of B –
probability of A and B
For example, say the probability of Marcie passing a test is 70%,
and the probability of Jerome passing the same test is 30%. The
events in this problem are independent: Neither person passing
the test influences whether the other person does so (unless of
course they cheat from each other, which we’ll assume is not the
case). So the final term of the expression, the probability of
Marcie and Jerome passing, is equal to the product of the
individual probabilities of those events, as we’ve seen all along.
We’ll convert the percentages to fractions, since working with
those may be easier. Then we’ll plug ‘em into the formula:

The probability of Marcie or Jerome passing the test is therefore

equal to 79%.
In some cases, two events may be mutually exclusive, meaning
that the probability of both occurring is 0. For example, in
choosing a single dog from a kennel, the chances of choosing a
black Labrador and a white schnauzer are zero—you can’t have
both. If these represented the A and B elements of the “or”
formula, the final term would be 0, and you wouldn’t have to
subtract anything.
Sequences
A sequence is a list of numbers that follows a particular pattern. If
you get a sequence problem, you’ll probably be given at least one
of the terms in the sequence, along with the rule that defines the
pattern. You probably won’t have to figure out the pattern on your
own; that’s more like the kind of thing you’d see on an IQ test.
However, what could make sequence problems tough is the
notation. Each term in a sequence has the same variable, but each
has a different subscript. This subscript indicates a particular
term. For example:
a1 = the first term
a2 = the second term
a10 = the tenth term
a n = the nth term
an+1 = the term immediately after the nth term
For example, to indicate that the second term of a sequence is 5
and that the third term is 7, the test makers might write:
a2 = 5
a3 = 7
This subscript notation can also be used to indicate how each term
relates to the others. For example:
an+1 = an – 3
This just means that each successive term is three less than the
previous term. Here’s an example, using the notation we just
discussed:
If a n = the nth term in a sequence, and a1 = 3 and an+1 = a n + 2,
what is the value of a10?
Let’s use 1 as n to keep things simple:
 a1 = a n = 3
So an+1 is the same as saying a1+1 or a2. This second term we’re told
is equal to the first term, a n , plus 2, which means that the second
term will be 3 + 2, or 5. So the notation, which looks intimidating,
is really a shorthand way of saying that each successive term is
two more than the previous term. Writing out this sequence from
the first to the tenth terms gives 3, 5, 7, 9, 11, 13, 15, 17, 19, 21. The
tenth term is 21, and so a10 = 21.
Arithmetic Sequences
In an arithmetic sequence, the difference between each term and
the next is constant. This is the kind of sequence we saw in the
previous example. In addition to understanding the notation and
concepts for sequences, you should know the formula for
arithmetic sequences:
a n = a1 + (n – 1)d
where
a n = the nth term
a1 = the first term
d = the difference between consecutive terms
This formula is useful if you need to determine the value of some
very high term and don’t want to write down a long sequence of
numbers. In our previous example, the first term (a1) is 3 and the
difference between consecutive terms is 2. If you plug these
numbers into the equation above looking for a10, you’ll get the
same answer, 21, that we got earlier. In this example, it’s just as
easy to write out the terms. But to determine the 100th term in
that sequence, we’ll need to plug the numbers into the formula:
a100 = 3 + (100 – 1)2 = 3 + 99 × 2 = 201
Geometric Sequences and Exponential Growth
In a geometric sequence, the ratio between one term and the next
is constant, not the actual difference between the terms. For
example, in the sequence 3, 9, 27, 81, each successive term is three
times greater than the preceding one, but the actual difference
between the terms changes: 9 – 3 = 6, 27 – 9 = 18, and so on.
Geometric sequences exhibit exponential growth, as opposed to
the constant growth of arithmetic sequences. Here’s an example of
the kind of geometric sequence that the test makers might toss at
you:
g1 = 4
g n = 2gn – 1
Trying out some terms, this means that g2 = (2)g1, g3 = (2)g2, and
so on. In other words, the first term is 4, and each successive term
is twice the value of the preceding term. Writing out the first few
terms lets us see that the ratio between terms is constant and thus
confirms that this is a geometric sequence:
4, 8, 16, 32, . . .
The ratio between consecutive terms is always 2, even though the
differences between the terms increase as you move to the right.
As with arithmetic sequences, you should learn the special
formula for geometric sequences, just in case it’s not convenient
to list out all of the terms up to the one you’re looking for:
g n = g1rn–1
where
g n = the nth term
g1 = the first term
r = the ratio between consecutive terms
Let’s use the formula to calculate the value of the tenth term in the
geometric sequence defined by g1 = 4 and g n = 2gn – 1. We already
know r = 2:
g10 = 4 × (2)10–1 = 4 × 29 = 4 × 512 = 2,048
Digit Counting
Here’s an interesting kind of problem that appears with some
regularity on the GRE. In digit counting problems, you’re asked
how many times a particular digit appears in a defined group of
numbers, or how many numbers within such a group don’t contain a
certain digit. An example will make this clearer:

How many three-digit positive even integers contain at least

one digit that is a 7?
 (A) 45
(B) 60
(C) 90
(D) 140
(E) 210
First make sure you understand the range of numbers under
consideration. Positive three-digit numbers begin with 100 and go
to 999. Moreover, we’re only interested in the even ones. The
question is looking for how many numbers fit this description and
contain at least one 7.
Well, we’re not going to go and count them all—that would take
too long. But we willlist a few examples that fit the criteria and
then see if we can discern a pattern. Beginning with the 100s, the
first number that has a 7 is 107, but since that’s not an even
number, it doesn’t count. In fact, the next one, 117, doesn’t count
either, nor does 127, 137, and so on. So we begin to see a pattern.
The first even number we get to that has a 7 is 170, followed by
172, 174, 176, and 178. No number from 179 to 200 fits the bill, so
in the entire 100s we have a total of 5 numbers that satisfy the
question’s requirements.
Now we can generalize from what we’ve learned: The 200s will be
no different, nor the 300s, nor the 400s, and so on. So if there are
5 cases in each of the nine groups of 100 numbers from 100 to
999, we can multiply 5 instances per each group of hundreds by
nine groups of hundreds (100s, 200s, etc.) to get 45, choice A.
However, we’d be wrong. Naturally, A is a trap for people who
forget that the 700s contain plenty of numbers that work, so we
need to consider the numbers from 700 to 800 separately. The
other eight groups of hundreds follow our pattern, so we’ll go with
5 × 8 = 40 to represent the instances that make the cut among
those. But we also have to consider all of the numbers in the 700s,
since every number in the 700s contains at least one 7. How many
are even? There are 100 numbers between 700 and 799, and half
are even, so we need to add 50 more cases to the 40 we’ve found
already. The correct answer is 40 + 50 = 90, choice C.
Digit counting questions can be difficult, and if you see one,
chances are the GRE software is throwing you more difficult
questions because you’re doing pretty well. Understand the range,
find a pattern and generalize it to as many cases as you can, and
then check for special cases. If you’re careful, you’ll find them all.
Factorials
You may see some problems on your exam in which a number is
followed by an exclamation point, like this: 5!. This does not mean
you should loudly exclaim, “Five!” Nor does it mean that the test
makers are extra-specially enthusiastic about a particular
problem. The exclamation point makes it look like there’s
something really exciting going on! But there’s not.
An exclamation point used in math symbolizes a factorial. A
factorial stands for the product of all the numbers up to and
including the given number. So 5! = 5 × 4 × 3 × 2 × 1 = 120.
Some more examples:
3! = 3 × 2 × 1 = 6
4! = 4 × 3 × 2 × 1 = 24
55! = 55 × 54 × 53 × . . . × 3 × 2 × 1 = a really huge number you
would never be expected to solve for
0! = 1
The proof of this last example is beyond the scope of what you
need to know for the GRE. Just remember that 0! = 1 by
definition. Consider it another bit of math trivia picked up on your
way to GRE mastery.
The factorial of n also signifies the number of ways that
the n elements of a group can be ordered. So, if you decide to
ditch grad school and become a wedding planner instead, and
need to figure out how many different ways six people can sit at a
table with six chairs, 6! is the way to go: 6 × 5 × 4 × 3 × 2 × 1 =
720 possible seating arrangements. You’ll astound the other
wedding planners with this quick calculation, never revealing the
true source of your knowledge.
As you might guess from the name, factorials have many factors.
Recall that a factor is a number that divides into another number
with no remainder. Whenever you take the factorial of a number,
the result will be divisible by all of the integers up to and
including the original number. For example, 6! is divisible by 6, 5,
4, 3, 2, and 1, and all of those numbers are factors of 6!. This is all
inherent in the definition of factorial, but it’s good to understand
it in these terms too.
Simplifying Factorials
The test makers may ask you to work out a problem that involves
factorials in fractions, and as you’ll soon see, this becomes
downright necessary in permutation and combination problems.
The trick is to cancel before calculating. As you’ve seen in earlier
examples, canceling with fractions means dividing the numerator
and denominator by the same number. A little cancellation makes
complicated-looking factorial problems much easier to solve.
Check it out:
What is  ?
This expression looks like it might be a huge number. And, in fact,
trying to calculate 98! or 100! would be near impossible without a
computer or ultra-fancy calculator. Fortunately, we can simplify
this equation significantly:

This works, because everything after and including the 98 cancels

out in both the numerator and the denominator, leaving 100 × 99
in the numerator and 1 in the denominator. Here’s another way to
think about this:

The 98! in the numerator cancels out with the 98! in the
denominator, leaving only 100 × 99 = 9,900.
Truth be told, even factorial problems involving smaller numbers
benefit from canceling out. For example:
Before you get sucked into multiplying out the factorials in the top
and bottom parts of the fraction and then dividing the results,
first cancel out what you can.
Permutations and Combinations
Knowing how to calculate and simplify factorials is especially
useful for problems involving permutations and combinations.
These types of questions ask you to determine how many ways
something can be done. They’re similar to the simple wedding
planner example above, except that they involve choosing a
certain number of entities from a larger group. For example, “In a
race of eight horses, how many ways can the horses finish first,
second, and third?” and “How many ways can two students be
selected for a Grammar Jamboree out of a class of 20 students?”
In this section, we’ll explain not only how to answer both of these
questions but also the very important difference between them.
Permutations
In a permutation, order matters—that is, being first in a group is
different from being second, third, or in any other position. The
easiest way to tell that a question is a permutation is if it includes
the word order or the word arrange. Even if it doesn’t contain
these words, the question might describe some kind of ranking or
race. Our horse race question above, for instance, is a permutation
since finishing first is certainly different from finishing second or
third in a horse race. If, for example, three of the horses are A, B,
and C, then ABC is one possible finish, CAB is another, and BCA is
a third. They all involve the same three horses, but switching them
around yields additional arrangements that we need to add to our
tally. In a combination problem, however, we’re not concerned
with order, so BCA would be considered the same as CAB, and we
wouldn’t count those twice. We’ll get to combinations in just a bit,
but let’s continue with the permutation problem at hand.
Here’s the permutation formula:

In this case, nPr is the number of subgroups of size r that can be

taken from within a set with n elements.
Here’s the horse race question again, and this time we’ll work
through it using the formula:
In a race of eight horses, how many ways can the horses finish
first, second, and third?
In this problem, n = 8, because that’s the total number of horses
racing, and r = 3, because that’s the number of winners we’re
interested in (first, second, and third place). Plugging into the
formula gives:

Now we’ll solve, using our knowledge of factorials and canceling

out:

Notice how the 5!s cancel out, leaving us with some basic
multiplication to get our final answer.
Combinations
As we mentioned earlier, in a combination, order does not
matter. For example, if you’re trying to buy three horses instead
of ordering them first through third in a race, then it really doesn’t
matter if you come away from the horse farm with horses ABC or
horses CBA—they’re all the same horses. You wouldn’t shuffle
them around and then say, “Look, a whole new group of horses!
Lucky me!” Well, you could say that, but people would think
you’re nuts.
You’ll be able to recognize a combination problem because it will
involve selecting a small group from a larger group, with no
regard to order. An example is the Grammar Jamboree problem
introduced earlier:
How many ways can two students be selected for a Grammar
Jamboree out of a class of 20 students?
Two students are to be selected from a class of 20, and no
mention of order is made. This is common for situations involving
teams: A team consisting of Jonathan and Gloria is the same as a
team consisting of Gloria and Jonathan. Unless some specific
mention is made of an ordering element, assume you’re dealing
with a combination problem. Here’s the combination formula:

Here, unordered subgroups of size r are selected from a set of

size n. Notice that this is the same as the permutation formula,
except that it tacks on an extra r! term in the denominator. This
means that we divide by a larger number in combination
problems, resulting in a smaller number of final orderings. And
that makes sense too: We’d expect fewer total orderings in
combinations, since order doesn’t matter and we therefore count
the shuffled orderings (ABC, CAB, BCA, etc.) as one.
Use the formula to solve our Grammar Jamboree problem:

Cancel the 18!s:

So, there are 190 possible two-person teams of jamboree-ers to

choose from the class of 20. Let the jamboree begin!
Multiple Permutations and Combinations
If the CAT software is really impressed by your math acumen, it
might throw you a problem involving multiple permutations,
combinations, or both. The key is to break the problem down into
parts, solving each independently using the formulas discussed
above. To obtain a final answer, multiply all of the individual
results. Here’s an example:
How many ways can Suzanne order an ice cream cone if she is to
select three different flavors out of fifteen available flavors and
three different toppings out of five available toppings? The order
in which the flavors are stacked is significant, but the order in
which the toppings are added is not.
We admit it—this looks hard. However, it helps to think of this as
two completely separate problems:
1. Selecting ice cream flavors
 2. Selecting toppings

Don’t worry, we’ll combine them in the end, much as the flavors
and toppings will be combined in Suzanne’s cone. But first things
first.
Since the order of flavors matters, this part of the problem is a
permutation. Suzanne must select three different flavors out of a
total of fifteen. Let’s use the P formula:

As per the problem, n = 15 and r = 3:

So Suzanne has 2,730 choices of flavors. Delicious! But back to the

problem: Since the order of toppings does not matter, this part is
a combination. Suzanne must select three different toppings out
of a total of five. Let’s use the C formula:

Now n = 5 and r = 3:

Cancel the 3!s, and restate 2! as 2 × 1:

Now we know that there are 10 possible combinations of toppings.

The last step is to multiply the two results: 2,730 × 10 = 27,300.
And there we go: Suzanne has an unbelievable 27,300 choices for
her ice cream cone. It’s a wonder people don’t have anxiety
attacks in ice cream stores—just one cone but more than 27,000
varieties to choose from!
Groups
This topic sounds innocuous enough, but problems involving
groups can drive test takers batty. The basic idea is that some
people or things belong to one group, others belong to another
group, and still others belong to both groups or neither group. For
example, at a certain country club, some members play golf, some
play tennis, others play both sports, and still others prefer reading
to playing. You’ll be given some of the specific numbers in such a
problem and then asked to determine the missing values.
The approach you should use depends on whether the problem
concerns two or three groups. We’ll cover the most effective
techniques for both cases. You won’t see problems with more than
three groups.
Problems with Two Groups
All you need for two-group problems is this formula:
group 1 + group 2 – both + neither = total
where
group 1 = the number of entities in one of the two groups
group 2 = the number of entities in the other group
both = the number of entities in both groups
neither = the number of entities in neither group
total = the total number of entities
You’ll most likely be given values for all of the parts of this
formula except for one. You’ll then have to determine the value of
the missing part. Let’s see how this works in the following
example.
At a certain animal refuge, 180 animals have four legs, 240 are
warm-blooded, and 85 both have four legs and are warm-
blooded. If the animal refuge has 500 animals in total, how many
animals at the refuge have neither four legs nor are warm-
blooded?
If you don’t know the formula above, this question could be a
nightmare. But the formula makes it very doable. We’ll let group
1 be the 180 animals that have four legs and group 2 be the 240
animals that are warm-blooded. Since we’re also given the
number of animals that belong to both groups and
the total number of animals, the missing value is the number of
animals that belong to neither group. Not surprisingly, that’s what
the question is after. Plugging the values into the formula gives:
180 + 240 – 85 + neither = 500
Solving for neither is a simple matter of solving this linear
equation with one variable, something we discussed way back in
the algebra section. Here goes:
180 + 240 – 85 + n = 500
335 + n = 500
n = 165
Voila! (That’s genuine excitement—not a factorial.) But things get
a bit more difficult if they throw in an extra group.
Problems with Three Groups
The formula for three-group problems is really long and
complicated (it involves nine unique terms!). So, we’ll skip it in
favor of an easier approach: Venn diagrams. Remember these
from junior high? Venn diagrams consist of intersecting circles, in
which each circle represents the number of entities in a particular
group. For example:

You’ll notice that the circles overlap. The upside-down triangular

section in the middle, with the darker shading, represents the
number of entities that belongs to all three groups. Sections in
which only two circles overlap, indicated with the lighter shading,
represent the number of entities that belongs to two overlapping
groups. The outermost section of each circle, the part that doesn’t
overlap with any of the other circles, represents the number of
entities that belongs to each group alone. For example, in the
swimming circle, the outermost section represents the number of
swimmers who neither lift weights nor do aerobics.
The key to three-group problems is to work from the inside out.
Begin with the entities that belong to all three groups, then
address the entities that belong to two groups, and finally deal
with the entities that belong to only one group.
Here’s an example of a three-group problem that conveniently
makes use of the diagram above:
At the Get Fit Athletic Club, every member swims, lifts weights,
does aerobics, or participates in some combination of these three
activities. Sixty members swim, 75 lift weights, and 100 do
aerobics. If 34 members both lift weights and swim, 25 members
both do aerobics and swim, 44 members both use the weight
room and do aerobics, and 10 participate in all three activities,
how many members belong to the Get Fit Athletic Club?
We’ll start by filling in values, working from the inside out. Since
10 people belong to all three groups, we’ll write 10 in the middle
section:

Next, we’ll fill in the values for people who belong to two groups.
We’re told that 25 members participate in aerobics and in
swimming, and it would be very tempting to write 25 in the
section above the 10. Keep in mind, however, that 10 of these 25
people have already been accounted for: The 10 people who
participate in all three activities are included among those who
participate in aerobics and swimming. That leaves 25 – 10 = 15
people for the aerobics/swimming overlap section above the 10 in
the middle. Similarly, 44 people do aerobics and use the weight
room. Since 10 of these people have already been accounted for in
the middle section, that leaves 44 – 10 = 34 people in the section
that overlaps aerobics and weight room. Also, 34 people swim and
use the weight room, so we’ll write 34 – 10 = 24 in the section
overlapping those categories. That brings us to here:
Next we need to fill in values for people who belong to only one
group. As with the previous step, we have to be very careful not to
count anyone more than once. For example, we’re told that 60
members swim. That means that the total of all the numbers in
the swimming circle must be 60. We already have 10 + 15 + 24 =
49 members in the swimming circle. That leaves 60 – 49 = 11
members for the outermost section of the swimming circle.
Similarly, since 75 members use the weight room, that leaves 75 –
10 – 34 – 24 = 7 members for the outermost section of the weight
room circle. Finally, since the total of the aerobics circle must be
100, the number in the remaining section must be 100 – 10 – 34 –
15 = 41. Filling these numbers in the appropriate sections of our
Venn diagram yields the following:

The problem asks for the total number of club members, and
we’re told specifically that every member participates in at least
one of these activities. That means that no member exists outside
of our three circles. We can therefore add all of the values in our
diagram to arrive at the total number of club members. This gives
us a final answer of 41 + 15 + 11 + 34 + 10 + 24 + 7 = 142.
We realize that this was a bit of work, but Venn diagrams are the
only way to go on three-group problems. And the good news is
that you’d have to be doing something right on the Math section
for the CAT software to throw something this nasty your way,
since the questions increase in difficulty the better you do.
That finishes up our discussion of data analysis. Keep in mind
that the GRE also requires you to analyze data presented in the
form of graphs and charts, but we have a whole chapter called
Data Interpretation devoted to that. In fact, we have a special
chapter devoted specifically to each of the three question types
you’ll see. It’s high time we got to those, where you’ll learn how
the concepts covered here in Math 101 play out in GRE math
questions, as well as effective ways to go about tackling them.
Problem Solving is the first question type we’ll cover, and it’s up
next.