Probability distributions

A DRV is a value obtained by taking a measurement from an experiment

in the real world. A random variable must take a numerical value.

Eg the outcome of rolling a fair dice Eg the outcome of picking a lottery ball

A random variable’s probability distribution gives the full

set of possible values and their associated probabilities.

Eg the outcome of rolling a fair dice Eg the number of heads obtained

when tossing a fair coin 3 times
x 1 2 3 4 5 6 Possible combinations: HHH
1 1 1 1 1 1 HHT
P(X = x) 6 6 6 6 6 6
x 0 1 2 3 HTH
THH
1 3 3 1 HTT
P(X = x) 8 8 8 8
HTH
TTH
TTT
From the above, you can see that  p x   1 for a DRV
 Probability functions
A probability function P(X = x) for a DRV is an algebraic function
giving the probability of each value in the distribution taking place

 6x , x  1,2,3 x 1 2 3
Eg P ( X  x)   1 2 3
0, otherwise P(X=x)
6 6 6

You will be asked to find the value of a constant k that makes

a probability function valid, using the fact that Σ p (x )  1
 p x   1
Eg find the value of k
x 1 2 3 4
kx, x  1,2,3,4
P( X  x)   1 2 3 4
 0, otherwise P(X=x) k
10
2k
10 3k
10
4
10
k

  1  p x   1
 p x  1  10 k  1  k 
10 Now try Ex8A
 Cumulative Distribution Functions
The cumulative distribution function of a
DRV gives the cumulative probability of X
F ( x)  P( X  x0 )   p( x)
x  x0
being less than or equal to a certain value

 10x , x  1,2,3,4 x 1 2 3 4
Eg P( X  x)   1 2 3 4
 0, otherwise P(X=x)
10 10 10 10
Write down (a) F(2) (b) F(3.6)
1 2 3 4 6
F(2) = P(1) + P(2)    F(3.6) = F(3) = 1 – P(4)  1 
10 10 10 10 10

Eg F  X  
 x  k  , x  1,2,3 Find the probability distribution of X
8
3
F ( x max )  1 1 5 6 P 1  F (1) 
F (1)   4 Probability distribution:
8 8
F  3  1 P 2  F  2  F (1) 
1
25 7 8 x 1 2 3

3  k   1 F (2)   1 3 1 1
8
8 8 P 3   F  3   F (2)  P(X=x)
8 4 8 8
k 5 Now try Ex8B
 Mean and Variance of a DRV
You have already used rules for the mean and standard deviation of data:


 x The square of the standard deviation is known
n as the variance of a data set.

  x2  2
Variance   
2  x 2

 2
The more spread out a set of data,
the larger the variance will be.
n n
The average and spread of a DRV can be quantified using similar
calculations known as the expectation E(X) and variance Var(X)

Var ( X )  E ( X 2 )   E ( X )
2

E  X    x. p( x)
where  
E X 2   x 2 . p( x)
 6x , x  1,2,3
Eg P( X  x)  
0, otherwise E  X   1 61  2  31  3  21  7
3

x 1 2 3  
E X 2  12  61  22  31  3 2  21  6
1 1 1
Var ( X )  6   73  
2
P(X=x) 6 3 2
5
9 Now try Ex8D
 Discrete Uniform Distributions
1
A DRV defined over the interval {1,2,3,…,n} with p  x  
n
for all values of X is known as a discrete uniform distribution

Eg the outcome of rolling a fair dice x 1 2 3 4 5 6

1 1 1 1 1 1
P( X  x)  61 , x  1 to 6 P(X = x) 6 6 6 6 6 6

Plotting this as a graph: Because the probability of each outcome

is the same, calculations with discrete
uniform distributions are relatively simple

Eg P X  3   36  1
2

Eg P ( X  5)  P  X  6   1
6

Even the expectation and variance can be

reduced to a simple formula determined by n
The expectation of a discrete Using algebraic manipulation covered in FP1, it can
uniform distribution is be shown that for a discrete uniform distribution:

E X  
n 1
Var ( X ) 
 n  1 n  1
2 12
WB13 The random variable X represents the number on the uppermost face when
a fair die is thrown.
(a) Write down the name of the probability distribution of X.
(b) Calculate the mean and the variance of X.

a) the discrete uniform distribution with n = 6

6 1 7  5 35
b) E  X    3.5 Var  X   
2 12 12
WB12 The random variable X has the discrete uniform distribution
Given that E(X) = 5, 1
(a) show that n = 9. P(X = x) = , x = 1, 2, ..., n.
n
Find
(b) P(X < 7), (c) Var (X).

a) E X   5  n 1
2 5 b) P  X  7   P1  X  6  c) Var ( X )  108
12  20
3

n9  96  2
3 Now try Ex8F
 Proving that for p x   1
n , x={1, 2, 3, …, n}, Var ( X ) 
 n 1 n 1
12

Var  X    x 2 p ( x)   E  X  
2
By definition,

p x   E X    Var  X    x    
n 1 2 2
1 1 n 1
For any DUD: n 2 n 2

 Var  X     n 1 2
Remove factor of
1
n
1
n x 2
 4

Using FP1 rule: x 2

 n  n 1 2 n 1
6  Var  X    n 1 2 n 1
6   n 1 2
4

Writing as a single fraction  Var  X   2  n 1 2 n 1  3  n 1 2

12

Expanding and simplifying  Var  X   n 2 1

12

Factorising  Var  X    n 1 n 1

12
 ‘Coded’ DRVs
Recap: the DUD for n = 6 x 1 2 3 4 5 6
1 1 1 1 1 1
What is E(X)? E ( X )  3.5 p(x) 6 6 6 6 6 6

Consider this distribution…

y 2 4 6 8 10 12
What is E(Y)? E (Y )  7 1 1 1 1 1 1
p(y) 6 6 6 6 6 6
The expectation is twice as big as E(X)

Consider this distribution…

z 8 9 10 11 12 13
What is E(Z)? E ( Z )  10.5 1 1 1 1 1 1
p(z) 6 6 6 6 6 6
The expectation is 7 higher than E(X)

What is the connection between X and Y? Y  2X

What is the connection between X and Z? Y  X 7
 The previous slide gives an example of a more general fact:

If X is a DRV, then: E  aX  b   aE  X   b
Eg a DRV has E(X) = 5. Find: E(2X-1)  2 E  X   1  9

E(5-4X)  4 E  X   5  15

Consider a DRV with x x1 x2 x3

the following E  X    x. p( x )
probability distribution: P(X=x) p1 p2 p3
 x1 p1  x2 p2  x3 p3

Now consider a DRV x ax1  b ax2  b ax3  b

obtained by ‘scaling
up’ to aX+b P(X=x) p1 p2 p3

E  aX  b    ax1  b  p1   ax2  b  p2   ax3  b  p3

 a x1 p1  x2 p2  x3 p3   b p1  p2  p3 
 aE  X   b p1  p2  p3  1
 The result for Var(aX+b) can be obtained using a similar approach

Var  aX  b   a 2Var  X 

Eg a DRV has Var(X) = 3. Find: Var(2X-1)  22 Var  X   12

Var(5-4X)    4  Var  X   48
2

Why is Var(X) multiplied by a2? Why isn’t Var(X) affected by b?

Recap: the DUD for n = 6 Consider Y = 3X + 1

x 1 2 3 4 5 6 y 4 7 10 13 16 19
1 1 1 1 1 1 1 1 1 1 1 1
p(x) 6 6 6 6 6 6 p(y) 6 6 6 6 6 6

Var ( X )  12
35
The +1 doesn’t affect Var(Y) but the x3 does

To summarise: Var (Y )  3 2  Var  X   105

4

E  aX  b   aE  X   b
Var  aX  b   a 2Var  X  Now try Ex8E
 WB9 A discrete random variable X has a probability function as shown in the table
below, where a and b are constants.
x 0 1 2 3
Given that E(X) = 1.7,
(a) find the value of a and the value of b. P(X = x) 0.2 0.3 b a
Find
(b) P(0 < X < 1.5), (c) E(2X – 3).

(d) Show that Var(X) = 1.41. (e) Evaluate Var(2X – 3).

a) E  X    x. p( x ) c) E  aX  b   aE  X   b

 0  0.2  1 0.3  2b  3a  1.7 E  2 X  3   2 E  X   3  2  1.7  3  0.4

 2b  3a  1.4 1 
d) Var ( X )  E ( X 2 )  E ( X ) 2
 p x   1 where  
E X 2   x 2 . p( x)
 0.2  0.3  b  a  1  
E X 2  02  0.2  12  0.3  22  0.1 32  0.4  4.3
 b  a  0.5  2 Var ( X )  4.3  1.7 2  1.41
1  2   2  a  0.4
 b  0.1 e) Var  aX  b   a 2Var  X 
b)  P 1  0.3 Var  2 X  3   22Var  X   4  1.41  5.64
WB10 The discrete random variable X has probability function
k ( x 2  9), x  4, 5, 6
P(X = x) =  where k is a positive constant.
0, otherwise,
1
(a) Show that k = (b) Find E(X) and Var(X). (c) Find Var(2X – 3).
50

a) x 4 5 6 b) E  X    x. p ( x)
7k 16 k 27k
P(X=x) E  X   4  507  5  16
50  6  50
27
 5.4
Var ( X )  E ( X 2 )   E ( X )
2

 p x   1
1
where  
E X 2   x 2 . p( x)
 50k  1  k 
50  
E X 2  42  507  5 2  16
50  6  50  29.68
2 27

giving Var ( X )  29.68  5.4 2  0.52

x 4 5 6
c) Var  aX  b   a 2Var  X 
7 16 27
P(X=x)
50 50 50 Var  2 X  3   22Var  X 
 4  0.52  2.08
 WB11 The random variable X has probability function

kx, x  1, 2, 3,
P(X = x) =  where k is a constant.
k ( x  1), x  4, 5,
(a) Find the value of k. (b) Find the exact value of E(X).
(c) Show that, to 3sf, Var(X) = 1.47 (d) Find, to 1dp, Var(4 – 3X).

a) x 1 2 3 4 5 b) E  X    x. p ( x)

P(X=x) k 2k 3 k 5 k 6 k E  X   1 171  2  172  ...  5  176  17

64

c) Var ( X )  E ( X )  E ( X )
2
 2
 p x   1 where  
E X 2   x 2 . p( x)
 17k  1  k 
1
17
 
E X 2  12  171  22  172  ...  52  176  266
17

  17   1.474...
2
Var ( X )  266
17
64
giving
x 1 2 3 4 5 d) Var  aX  b   a 2Var  X 
1 2 3 5 6
P(X=x) Var  4  3 X   3 2Var  X 
17 17 17 17 17
 9  1.474...  13.3
WB13 The random variable X represents the number on the uppermost face when a
fair die is thrown.
Three fair dice are thrown and the numbers on the uppermost faces are recorded.
(c) Find the probability that all three numbers are 6.
(d) Write down all the different ways of scoring a total of 16 when the three numbers
are added together.
(e) Find the probability of scoring a total of 16.

c)  61  3  2161
d) Either two 5’s and a 6 or two 6’s and a 4, each with 3 combinations

5,5,6 5,6,5 6,5,5

6,6,4 6,4,6 4,6,6

e) Using (d), 6 outcomes out of 216 possible, each equally likely

6
216