Fundamentals of Probability
Fundamentals of Probability
Fundamentals of Probability
Outcomes of experiment
Elements
of Sample
Individual
outcome inSpace
Outcomes
of survey
experiment
statistical
etc.
Individual outcome in
0
Collectively those real numbers become;
For example, if we record last years income and other important characteristics of 30,000
randomly chosen households in Vietnam, then the random variables are;
X: income
Random variables are always defined to take on numerical values, even when they describe
qualitative events.
For example, consider housing ownership at the sample survey of 30,000 H/H above, where
the two outcomes are own house and rented house.
We can define a random variable as follows: Y = 1 in case of own house, and Y = 0 in case
of rented house
In practical sense, random variables that take on numerous values are best treated as
continuous.
The pdf of a continuous random variable is defined only to compute events involving
a range of values.
For example, if a and b are constants where a b, the probability that X lies between the
numbers a and b, P(a X b), is the area under the pdf between points a and b, as shown in
Figure B.2.
the integral of the function f between the points a and b.
The entire area under the pdf of a continuous random variable must always equal one.
10
11
12
13
Consider a basketball player shooting two free throws. Let X be the Bernoulli random variable equal to one if
she or he makes the first free throw successful, and zero otherwise. Let Y be a Bernoulli random variable
equal to one if he or she makes the second free throw success. Suppose that she or he is an 80% free throw
shooter,
so that P(X = 1) = P(Y = 1) = 0.8. What is the probability of the player making both free throws successful
If X and Y are independent, we can easily answer this question: P(X = 1,Y = 1) = P(X = 1) P(Y
= 1) = (.8)(.8) = 0.64. Thus, there is a 64% chance of making both free throws successful.
Independence of random variables is a very important concept.
If X and Y are independent and we define new random variables g(X) and h(Y) for any functions g and h,
then these new random variables are also independent.
If , , , are discrete random variables, then their joint pdf is
f (, , , ) = P( = , = , , = ).
The random variables , , , are independent random variables if, and only if, their
joint pdf is the product of the individual pdfs for any (, , , ). This definition of
independence also holds for continuous random variables.
Assuming statistical independence often provides a way to reasonable approximation in more
complicated situations.
14
15
If the flight has 100 available seats, the airline is interested in P(X > 100). Suppose, initially,
that n =120, so that the airline accepts 120 reservations, and the probability that each person
shows up is = 0.85. Then, P(X >100) = P(X = 101) + P(X = 102) + +P(X = 120), and each
of the probabilities in the sum can be found from equation (B.14) with n = 120, = 0.85, and
the appropriate value of x (101 to 120).
If n = 120, P(X >100) = 0.659
High risk of overbooking
If n = 110, P(X >100) = 0.024
Manageable risk
16
for all values of x such that (x) > 0. The interpretation of (B.15) is most easily seen when X
and Y are discrete. Then,
(yx) = P(Y = yX = x),
B.16
where the right-hand side is read as the probability that Y = y given that X = x. When Y is
continuous, (yx) is found by computing areas under the conditional pdf.
17
This means that the probability of the player making the second free throw depends on whether
the first free throw was made: if the first free throw is made, the chance of making the second
is .85; if the first free throw is missed, the chance of making the second is .70. This implies
that X and Y are not independent; they are dependent.
We can still compute P(X = 1,Y = 1) provided we know P(X = 1). Assume that the probability
of making the first free throw is .8, that is, P(X = 1) =.8. Then, from (B.15), we have
P(X = 1,Y = 1) = P(Y = 1|X = 1) P(X = 1) = (.85)(.8) = .68.
18
19
Given a random variable X and a function g(.), we can create a new random variable g(X).
X:is random variable.
and log(X) : are random variable
The expected value of g(X) is, again, simply a weighted average:
E[g(X)] =
B.19
or, for a continuous random variable,
E[g(X)] =
B.20
20
ExampleB.4
[Expected Value of ]
For the random variable in Example B.3, let g(X) = . Then,
E() = (1/8) + (1/2) + (3/8) = 13/8.
In Example B.3, we computed E(X) = 5/8, so that = 25/64.
E()
In fact, for a nonlinear function g(X), E[g(X)] g[E(X)] (except in very special cases).
If X and Y are random variables, then g(X,Y) is a random variable for any function g.
Expectation of joint distribution. When X and Y are both discrete, taking on
values {, , , } and {, , , }, respectively, the expected value is
E[g(X,Y)] =
where is the joint pdf of (X,Y).
21
22
23
24
25
26
27
Thus, the random variable Z has a mean of zero and a variance (and therefore a standard
deviation) equal to one
Ex.) Suppose that E(X) = 2, and Var(X) = 9. Then, Z = (X 2)/3 has expected value zero
and variance one.
29
30
Covariance
Let = E(X) and = E(Y) and consider the random variable (X) (Y).
Now, if X is above its mean and Y is above its mean, then (X) (Y) 0. This is also
true if X and Y
On the other hand, if (X) and (Y) , or vice versa, then (X) (Y) 0.
The covariance between two random variables X and Y is defined as;
Cov(X,Y) = E[(X) (Y)],
B.26
which is sometimes denoted .
If 0, then, on average, when X is above its mean, Y is also above its mean.
If 0, then, on average, when X is above its mean, Y is below its mean.
Useful expressions for computing Cov(X,Y) are as follows:
Cov(X,Y) = E[(X) (Y)] = E[(X )Y]
= E[X(Y )] = E(XY) .
B.27
It follows from (B.27), that if E(X) = 0 or E(Y) = 0, then Cov(X,Y) = E(XY).
31
32
33
34
Note) The variance of the difference is the sum of the variances, not the difference in the
variances.
Ex.) Let X denote profits earned by a restaurant during a Friday night and let Y be profits
earned on the following Saturday night. Then, Z = X + Y is profits for the two nights.
Suppose X and Y each have an expected value of $300 and a standard deviation of $15
(so that the variance is 225).
Expected profits for the two nights is E(Z) = E(X) + E(Y) = 2(300) = 600 dollars. If X and
Y are independent, and therefore uncorrelated, then the variance of total profits is the sum
of the variances: Var(Z) = Var(X) + Var(Y) = 2(225) = 450. It follows that the standard
deviation of total profits is or about $21.21.
36
When Y is continuous, E(Y|x) is defined by integrating y(yx) over all possible values of y.
The conditional expectation is a weighted average of possible values of Y, but now the
weights reflect the fact that X has taken on a specific value.
E(Y|x) is just some function of x, showing how the expected value of Y varies with x.
Ex.), Let X is years of education and Y is hourly wage.
Then, E(Y|X = 12) is the average hourly wage for working population with 12 years of
education (roughly a high school education).
Tracing out the expected value for various levels of education provides important
information on how wages and education are related. See Figure B.5 for an illustration.
In econometrics the relation between average wage and amount of education is specified
by simple function, we typically specify simple functions that capture this relationship. As
an example, suppose that the expected value of WAGE given EDUC is the linear function
E(WAGE|EDUC) = 1.05 + .45 EDUC.
39
40
41
Ex.) Let Y = SAVING and X = INCOME (both of these measured annually for the
population of all families). Suppose that Var(SAVINGINCOME) = 400 + .25INCOME.
This says that, as income increases, the variance in saving levels also increases. It is
important to see that the relationship between the variance of SAVING and INCOME is
totally separate from that between the expected value of SAVING and INCOME.
Property CV.1: If X and Y are independent, then Var(YX) = Var(Y).
This property is pretty clear, since the distribution of Y given X does not depend on X,
and Var(YX) is just one feature of this distribution.
43
45
47
48
Other distributions, such as income distributions, do not appear to follow the normal
probability function. In most countries, income is not symmetrically distributed
about any value;
the distribution is skewed toward the upper tail.
In some cases, a variable can be transformed to achieve normality. A popular
transformation is the natural log, which makes sense for positive random
variables.
If X is a positive random variable, such as income, and Y = log(X) has a normal
distribution, then we say that X has a lognormal distribution. It turns out that the
lognormal distribution fits income distribution pretty well in many countries.
Other variables, such as prices of goods, appear to be well described as
49
log-0normally distributed.
No simple formula can be used to obtain the values of (z) [because (z) is the integral
of the function in (B.35), and this integral has no closed form].
The values for (z) are easily tabulated; they are given for z between 3.1 and 3.1.
For z 3.1, (z) is less than .001, and for z 3.1, (z) is greater than .999.
50
The df in a chi-square distribution corresponds to the number of terms in the sum in (B.41).
Have an important role in statistical and econometric analyses.
From equation (B.41), a chi-square random variable is always non negative.
Also, the chi-square distribution is not symmetric about any point.
If X ~ , then the expected value of X is n [the number of terms in (B.41)], and the
variance of X is 2n.
51
53
54
55
56
Practice 2.
Calculation of Covariance and Correlation Coefficients Using SPSS
1. Choose 3 pairs of indicators for which you think that they are highly correlated from
Socio- economic Indicators of Vietnam (Monthly or Quarterly).
2. Again choose 2 pairs of indicators for which you think that they are not correlated.
3. Calculate the covariance and correlation coefficients for each of 5 pairs of indicators.
4. Now produce = - for each of indicators you have and calculate the
correlation coefficients.
5. Compare the correlation coefficients from original data to those from difference.
57
K EYTE R M S
Bernoulli (or Binary)/Random Variable / Binomial Distribution / Chi-Square Distribution
Conditional Distribution / Conditional Expectation / Continuous Random / Variable
Correlation Coefficient / Covariance / Cumulative Distribution / Function (cdf) /
Degrees of Freedom / Discrete Random Variable / Expected Value Experiment /
F Distribution Independent Random Variables / Joint Distribution / Kurtosis /
Law of Iterated Expectations / Median / Normal Distribution /
Pairwise Uncorrelated Random Variables / Probability Density Function (pdf) /
Random Variable / Skewness / Standard Deviation / Standard Normal Distribution /
Standardized Random Variable / Symmetric Distribution / t Distribution /
Uncorrelated Random Variables / Variance
58