You are on page 1of 7

Hello everyone!

I hope you had a good day so far and are ready for a good and
informative seminar tonight! We have a lot to talk about today.

The Unit examples and some odd-numbered problems that I posted on doc sharing should
fairly help you with understanding of this Unit’s concepts once you read the chapters and
my seminar notes after the seminar.

I know this class is very busy and challenging. Just remember that I am here to help. If
you can, please try to come to the office hours on Sundays AIM- 9:00 pm to 11:00 am ET
or post your questions in each Unit to get help. Also, as I said before, you can use the
NetTutor which is the online tutoring help for this course. The link to that site is on your
MyDesk page. I really want all of you succeed in this class.

Ok, I am going to help you with your amount of reading a little this week. You can skip
section 5.6 (Bayes), Poisson distribution in Chapter 6 if you haven't read them seriously
yet.

You had a lot to learn this week so I guess it would be fair to skip these sections that you
will not need in this course later on. There isn't any problem from these sections in your
assignments either.

I know it is not easy to learn everything in the seminar in such a short amount of time
especially if you have not had a chance to read the chapters before the seminar time. Just
make sure to go back to the seminar archive and read the part you missed or didn't fully
understand in the seminar. I try to clarify as much as I can but if you still had questions
please just ask.

I am going to share some easy to follow examples with you (they are not in the book) that
helps you to understand these concepts even if you have not had a chance to read the
chapters 5 and 6 yet.

Let us start this seminar discussion with flipping a coin once (chapter 5). Theoretically,
the probability of head showing is 0.5. Now, if I want to find the probability of head on
the second flip when the first flip was a head, the answer is still 0.5 because the first flip
and second flip are independent of each other. In other words what happens on the second
flip doesn't depend on what happened on the first flip.

So if we call the first head showing -event A- and second head showing -event B- then
p(B|A) = P(B) = 0.5 where “p” means probability. So, P(B) means probability that event
B happens. P(B|A) means probability that event B happens condition (if) event A
already happened (section 5.2).
Now, let us say A and B are two independent events. Event A is heads showing on the
first flip and event B is heads showing on the second flip. So, when flipping a coin twice,
we can say that probability of heads showing on the first time AND on the second time is
p (A and B) = P(A) times p (B) = 0.5 * 0.5 = 0.25. So, whenever you have the word
AND you have to multiply probabilities here (section 5.5).

Now, let us work on another example. Let us say you have 3 red marble balls and 2 blue
marble balls in a hat. You draw one ball at random. Can someone tells me what the
probability of getting a red ball is?

yes, it is 3/5 three red out of total of 5 balls. Is it clear for everyone why it is 3/5?

Now, what is the probability that the second draw is a red condition that the first draw
was a red?

Let us call getting red on first try is event A and getting red on second draw is event B.
The answer is 2/4 or 1/2 or 0.5.

one red is out already and total is reduced to 4 so it is 2/4 or 1/2.

so we have P(B| A) = P(B condition A happened) = 1/2.

So, what the “condition” part of the problem do for us is reducing the sample stace for us
because knowing that one red is already out reduces the total sample space to 4 and
knowing that it was a red reduces the total remaining red marbles to 2 (from original 3).

Is it clear so far?

Now, lets talk about something interesting and related to this subject. It is good to try to
"visualize” the chance of winning a lottery. Imagine a book as big as ours with just all
zeros in each page. Your chance of hitting the jackpot is one of the zeros in that book!

Are you ready for another fun example :)

You see an exclamation mark, “!”, is used in Binomial Distribution formula. Exclamation
mark is called Factorial in Math and in Statistics. For instance, 3! = 3*2*1= 6. 5! =
5*4*3*2*1 = 120. So, to evaluate a factorial you start multiplying the number to all the
numbers less that it all the way to the number 1.

Here is another example. (5-2)! = 3! = 3*2*1 = 6.


And here is another one. 8! / 6! = 8*7*6*5*4*3*2*1 / 6*5*4*3*2*1 = 8*7 = 56 (we
cancelled out all values from 2 through 6 from the numerator and the denominator)

We can do it in a different way that is more efficient. Here it is:

8! / 6! = 8*7 *6! / 6! = 8*7 = 56

Any questions about factorial symbol now?

Book talks about permutation and combination. Here is an explanation and example for
both.

In permutation ORDER IS IMPORTANT. Lets say we have 3 letters A, B, and C. How


many ways I can select group of 2 letters out of these 3 letters? Well, they are AB, BA,
AC, CA, BC, and CB. So there are 6 different ways.

So, you can use =PERMUT(3,2) to get the number of ways to select a subgroup of 2 out
of 3 total elements when order is important.

The answer would be 3! / (3-2)! = 3! / 1! = (3*2*1) / 1 = 6

Now, in Combination situation, THE ORDER IS NOT IMPORTANT. In other words, if


we are looking for possible ways of getting groups of 2 letters out of 3 letters A, B, and C
and as long as we have A and B in a group the order of which one comes first is not
important we use Combination formula.

So, in this case, we use AB, BC, and AC because CA, BC, and BA are redundant.

So, you can use =COMBIN(3,2) to get the number of ways to select a subgroup of 2 out
of 3 total elements when order is NOT important.

The answer would be 3! / 2! (3-2)! = 3! / 2! 1! = (3*2*1) / 2 = 3

Now, lets talk about Binomial distribution. The Computer Solutions 6.1 tells us how to
use it. I am going to use it here in a few different examples so you can understand how to
apply this formula in a situation.

We can apply Binomial probability to solve problems in 3 different wasy. BINOMDIST


Excel formula, Computer Solution 6.1 way, and Table in the back of the book. I will give
you all one example from each tonight.
BINOMDIST(r value, total trial, probability of success, cumulative) is the Excel
command that we can use to get the value of a Binomial problem for us. In this
command, r is the number of success we are looking for and cumulative is either TRUE
or FALSE. . Just click on a cell and type the command in the text box of Excel.

I am giving you the following example to show the effect of parameter cumulative in the
formula.

Example 1: What is the probability of selecting a group of 2 out of a total of 5 objects


when probability of success is 30% (0.3)? Here is the answer.

BINOMDIST(2, 5,0.3,FALSE)= 0.3087 (the value is 0.309 on page 933 in the back of
the book)

As you see, if you use the parameter *False* then the formula really is finding probability
of getting 2 success out of 5 trials.

Example 2: What is the probability of selecting a group of 2 or LESS (or, at most 2) out
of a total of 5 objects when probability of success is 30% (0.3)? Here is the answer.

if you use Excel and enter the following command you get the answer of 0.83692. So
=BINOMDIST(2, 5,0.3,TRUE)= 0.83692.

This is the result of having 0 success, 1 success or 2 success out of total of 5. Again,
success can be anything. It can be being a female in the group, having a height of higher
than 6 feet, getting A on an exam, etc.)

You can also use in the back of the book page (page 933) to find the Binomial answers.
For this problem the value is also 0.837 on page 933in the back of the book. You cannot
find the value of 0.837 in the table directly. Let me tell you what is going on here.

Lets go to page 933 and look at n=5 section on that page. This is the total value. Look at
K=2 value. Just go to the right on K=2 and stop under the column of 0.3. You will see
the value of 0.3087. This is the same value that we got from our Excel formula. Here is
the formula again. BINOMDIST(2, 5,0.3,FALSE)= 0.3087

Now, look at K=1 value. Just go to the right on K=1 and stop under the column of 0.3.
You will see the value of 0.3601. This is the same value that we got from our Excel
formula. Here is the formula again. BINOMDIST(1, 5,0.3,FALSE)= 0.3601

K=0 value. Just go to the right on K=0 and stop under the column of 0.3. You will see
the value of 0.1681. This is the same value that we got from our Excel formula. Here is
the formula again. BINOMDIST(0, 5,0.3,FALSE)= 0.1681
Now, if you want to know the probability of 2 or less (or at most 2) successes you are
looking at adding k=0 and k=1 and k=2 values so it will be 0.3087+ 0.3601 + 0.1681 =
0.8369.

Any questions so far?

Now, I have a question? What if I want to find the probability of 2 or 3 success out of 5
trials?

BINOMDIST(2, 5,0.3,FALSE) + BINOMDIST(3, 5,0.3,FALSE)= 0.3087 + 0.132

You can also get the value of a binomial by using page 933 at the end of the book.

Let's do a little practice. What is the probability when n=5, r=1 and p= 0.3 from table 3?

Yes it is 0.360 or 36%.

Now, lets say, you play a game with your friend and say we flip a coin. If head comes up
you get $1 and if tail comes up I get a $1.

These discussions make a lot more sense if we go back to it and read the seminar archive
after the chat again.

now, what do you think happens if you all keep playing many times?

Good! you get even. they call it -fair game-. in a fair game, no one loses or wins. it is a
draw. Expected value = 1 (if you win) times 0.5 + -1 (if you lose) times 0.5 = 0.5 - 0.5 =
0. We use -1 to show the loss of $1 and use +1 to use gain of $1. The value of 0.5 is
probability of losing or gaining $1. It is 50/50 to lose or win.

Remember that “luck” affect the outcome in short periods. If you play for 10 minutes you
many lose a lot more or win a lot more times. We are talking about playing for a long
long time (theoretically).

Are you ready for another example?

Example of rolling a die. lets say you get $10 if you get number 2 showing and lose $3 if
you get anything else showing. Expected value is = 10 times 1/6 + (-3) times 5/6 = -5/6.
So, on the long run if you keep playing this game you will lose –5/6 dollar per hand.

Here, we have one chance to get 2 showing out of 6 numbers on a die so its probability is
1/6. Also chance of number 2 not showing is 5/6 because there are 5 numbers other than
2 that can come up.
Any questions here?

Let me give you an easy example for Binomial Distribution. You need to pay attention to
the next 4-5 statements to get the whole picture. Let’s say we flip a coin 3 times. What is
the probability that we get 2 tails? Well, let’s see what is going on here. I am going to
discuss the answer in the next 4-5 posts.

All of the possibilities of the result of flipping a coin three times are as follow: HHH,
HHT, HTH, HTT, THT, THH, TTH, TTT. For example, in case of HTH pattern, we mean
first flip be Heads, second flip be Tails, and third flip be Heads etc.

Now, the number of Tails that we might have in flipping a coin 3 times is 0 (when no
Tails), 1, 2, or 3 (when all are Tails). Now, to answer, what is the probability that we get 2
Tails out of flipping a coin three times, we use Binomial formula. In this case, the
formula is C3,2 p^2 (1-p)^1 (it is C3,2 mulitply by p to the power of 2 multiply by q to
the power of 1). The symbol ^ is used to show the power of the term.

Here we use p for the symbol -PI- that is used in our book. What is p in this problem? It
is probability of success which is probability of getting a Tail (getting Tails is the success
for us because we are interested in probability of getting 2 Tails). Probability of getting 2
Tails is p.p = p^2 and since the last one is obviously a Head, it is a failure and its
probability is q^1.

So, we now have to solve C3,2 p^2 q^1 (C3,2 multiply by p to the power of 2 multiply
by q to the power of 1) or more precisely, C3,2 (0.5^2) (0.5^1). As we know, probability
of Tails or Heads showing is 0.5.

Now, what C3,2 represents in the formula? It tells you the number of ways you can have
2 Tails in flipping a coin 3 times. Look at all possibilities again: HHH, HHT, HTH, HTT,
THT, THH, TTH, TTT.

If you count the number of patterns that give you 2 Tails you will see it is equal to 3
(HTT, THT, TTH). We can also verify that we have three ways to get 2 Tails by
expanding the C3,2 formula. C3,2 = 3! / [2! * (3-2)! ] = 3! / 2! * 1! = (3.2.1) / (2.1) = 3.
You may say why we need the formula if we can visually check that.

Well, what if we were flipping 150 times and were looking for 57 tails or heads? That is
where combination formula shines.

So, to answer the problem of what is the probability of getting 2 Tails in flipping a coin 3
times so far we have: 3 (0.5)^2 (0.5)^1. We just need to work it out to get the probability
of getting 2 Tails in flipping a coin 3 times. The answer is 0.375 or 37.5% chance of
getting 2 tails when flipping 3 coins.
Remember the value of 3 in the formula was the result of how many ways we could get 2
success out of 3 trial (and success was getting tails).

You might also like