Discuss the Behavior of Random

Variables
STATISTICAL
P ( X = k) - - - ?
ESTIMATION P(X>k)---?
P(a<X<b)---?
Use Probability Distribution
Functions

BINOMIAL DISTRIBUTION POISSON DISTRIBUTION

Probability Distribution Function Probability Distribution Function

fx(X) = P(X = x) = nCx (P)x (1 - P) n-x fx(X) = P(X = x) =

n--- ? P---? λ--- ?

 NORMAL DISTRIBUTION VALUES OF THESE PARAMETERS - - ?
Probability Distribution Function
ESTIMATE PARAMETERS
USING
SAMPLES
µ--- ? σ---?

STATISTICAL ESTIMATION

STATISTICAL ESTIMATION POINT ESTIMATION

• Population Mean µ is estimated by Sample
by by Mean


SINGLE RANGE • Population Variance σ2 is estimated by

Sample Variance S2
VALUE OF VALUES
• Population Proportion P is estimated by
Sample Proportion π
 INTERVAL ESTIMATION IF (4 , 5) is an interval
• (3 , 6) also an interval
Let (4 , 5) is an interval estimate • (2 , 7) also an interval
• (1 , 8) also an interval
Then population mean lies between 4 and 5
If there is a single interval which includes
Therefore it also lies between 3 and 6 population mean
( 3 , 6 ) is also an interval for the population We can define infinite number of intervals.
mean

CONFIDENCE INTERVAL CONFIDENCE INTERVALS

Interval (2 , 5) is embedded in the interval (1 , 6) • Larger the interval , Higher the confidence

Therefore we are more confident in saying that • Smaller the interval , Lower the confidence
interval (1 , 6) includes the parameter
than saying • Each interval is associated with a certain level
of confidence
the interval (2 , 5) includes it.
[ ( θ ) ]
1 2 5 6 • They are called confidence intervals.
 EXPECTED VALUE OF VARIANCE OF
Let Xi ∼ N ( µ , σ2 ) Let Xi ∼ N ( µ , σ2 )
E(Xi) = µ and V(Xi) = σ2 for all i E(Xi) = µ and V(Xi) = σ2 for all i
Then Then
E ( X1 + X2 + . . . . . . + Xn) = E(X1) + E(X2) + . . . . +E(Xn) V ( X1 + X2 + . . . . . + Xn) = V(X1) + V(X2) + . . . . +V(Xn)
= σ2 + σ2 + . . . .+ σ2 = n σ2
=µ+µ+....+µ=nµ
V ( ΣXi ) = nσ σ2
E ( ΣXi ) = n µ ∑
 nσσ2 σ2
∑
  V = 2 =
E = =µ  n n
σ 2
   ) =
V(

 ) = µ
E(
n

σ2 95% CONFIDENCE INTERVAL

E( )=µ V( )=
n
2.5% 2.5%
 ~ ( ,  /)

 − 


 = ~ ( , )
/ 

If Z ∼ N ( 0 , 1)

P ( - 1.96 < Z < 1.96 ) = 0.95

  
 −.   −  ≤ .  = . 
≤

 

 TO EACH PART OF THE INEQUALITY

ADD −

P ( - 1.96 < Z < 1.96 ) = 0.95
 
  


 − . 
 −
 + .  = . 
≤ −  ≤ −

 −.  < < .  = 0.95  
/ 
MULTIPLY THE WHOLE INEQUALITY BY / 
MULTIPLY THE WHOLE INEQUALITY BY (-1)
 
 −.   −  ≤ .  = . 
≤
 
   + . 

 − .  = . 
≥  ≥

 

 
 + . 

 − .  = . 
≥  ≥
• 95% CONFIDENCE INTERVAL FOR THE MEAN
   
 − . 

 + . 
,

REARRANGING THE INEQUALITY  
 
 − . 

 + .  = . 
≤  ≤

  • 90% CONFIDENCE INTERVAL FOR THE MEAN
 
 − . 

 + . 
,

95% CONFIDENCE INTERVAL FOR THE MEAN  
 
 − . 

 + . 
,

  • 99% CONFIDENCE INTERVAL FOR THE MEAN
 
 − . 

 + . 
,

 
 EXAMPLE
(P/2%
(P/2)% A sample of size 25 was selected from a
(1-P)%
population with variance 16. If the sample mean
is 42 obtain 95% confidence interval for µ.
-Z 0 Z

P Z P Z
95% CONFIDENCE INTERVAL FOR THE MEAN
90% 0.1257 20% 1.2816  
80% 0.2533 10% 1.6449  − . 

 + . 
,

70% 0.3853 5% 1.9600
60% 0.5244 2% 2.3263
 
50% 0.6745 1% 2.5758
40% 0.8416 0.2% 3.0902
30% 1.0364 0.1% 3.2905
 = 42

σ=4 n = 25

95% CONFIDENCE INTERVAL FOR THE RICE FILLING EXAMPLE

MEAN
Senaweera Rice Mills used a machine to fill rice in
 = 42

σ=4 n = 25 to different bags. Due to the adjustment
 weaknesses the average weight per bag differs from
 − . 
Lower Limit =

 bag to bag but the variation of the fillings per bag
= 42 – 1.96* 4/5 remains constant at 16g. In order to estimate the
= 40.432 mean filling per bag, a sample of 25 bags was
 + .   selected and the average weight was estimated as
Upper Limit =

 52kg. If the owner wants to determine the range
= 42 + 1.96* 4/5 for the mean filling per bag with 90% confidence,
= 43.568 how can you help him?
95% Confidence Interval = [40.432 , 43.568]
  = 52

σ=4 n = 25
95% CONFIDENCE INTERVAL FOR THE MEAN
 
 − . 

 + . 
,

  Upper Limit  + . 
=


 = 52

σ=4 n = 25 = 52 + 1.64* 4/5
= 53.312

Lower Limit  − . 
=


90% confidence Interval for the mean is
= 52 – 1.64* 4/5
[50.688 , 53.312]
= 50.688

UNKNOWN VARIANCE σ 2
• ESTIMATE BY SAMPLE VARIANCE S2
 − 

• ( n – 1 ) – Degrees of Freedom
• THEN THE STATISTIC BECOMES
!/  • Number of independent square terms used in
 − 


the calculation of S2
HAS ONE SOURCE OF VARIANCE,

/  
∑


• S2 = - n square terms

  


&S
HAS TWO SOURCES OF VARIANCE,

!/ 
  
 - one term depend on other terms
•



t= ∼ tn-1 (student t distribution) • Only (n-1) independent terms
!/ 
 When n = 25
α/2)%
(α
Critical value for 95% confidence interval,
α/2)%
(α α)%
(1-α
t = 2.064
95% confidence Interval
-t 0 t " "
 − . 

 + . 
,

Degrees of Probability α  
Freedom 0.2 0.1 0.05 0.02 0.01

1 3.078 6.314 12.706 31.821 63.657 When n = 20

2 1.886 2.920 4.303 6.965 9.925
3 1.638 2.353 3.182 4.541 5.841 Critical value for 90% confidence interval,
4 1.533 2.132 2.776 3.747 4.604
5 1.476 2.015 2.571 3.365 4.032 t = 1.729
6 1.440 1.943 2.447 3.143 3.707
7 1.415 1.895 2.365 2.998 3.499 95% confidence Interval
8 1.397 1.860 2.306 2.896 3.355 " "
 − . # 

 + . # 
,

9 1.383 1.833 2.262 2.821 3.250  
10 1.372 1.812 2.228 2.764 3.169

Sample mean = 3789.55

MORTGAGE PAYMENT EXAMPLE Sample variance = 1270.762
t19,0.05 = 2.093 n= 20
Newly elected Mayor is interested in identifying the
monthly mortgage payments of all the citizens in  − . $ "
his Municipal Council area. For this purpose he
• Lower Limit =



select a sample of 20 households randomly and the = 3789.55 - 2.093* 1270.76/ 

following information on their monthly payments
= 3194.82
(Rs.) were collected.
"
• Upper Limit =  + . $




2563 3482 1752 6288 4575
= 3789.55 + 2.093* 1270.76/ 
3963 5463 3962 3500 4578
2785 4625 5600 2875 3967 = 4384.28
1595 4255 3762 1945 4256 95% confidence interval for µ = [3194.82, 4384.28]
 NECESSARY INSTRUCTIONS
MICROSOFT EXCEL OUTPUT
1. Select Data tab
Column1
2. Select Data Analysis
Mean 3789.55 3. Select Descriptive Statistics
Standard Error 284.1500111
Median 3962.5 4. Specify the Data Range
Mode #N/A
5. Define Output Location
Standard Deviation 1270.757481
Sample Variance 1614824.576 6. Select Summary Statistics
Kurtosis -0.355987652
Skewness -0.028175655
7. Click OK
Range 4693
Minimum 1595 Lower Limit = 3789.55 – 594.73 = 3194.82
Maximum 6288
Sum 75791
Count 20 Upper Limit = 3789.55 + 594.73 = 4384.28
Confidence Level(95.0%) 594.732807

MINITAB OUTPUT TELEVISION PICTURE TUBE EXAMPLE

One-Sample T: Morgage A small electronics firm claims that it has developed a new
Variable N Mean StDev SE Mean 95% CI cheap process for manufacturing colour television picture
Morgage 20 3789.55 1270.76 284.15 (3194.82, 4384.28) tubes. Assume they approached Mr. Siridasa and request
him to invest in their company so they can begin mass
production of the picture tubes. But what Mr. Siridasa
NECESSARY COMMANDS wanted to know is that how durable the picture tubes
manufactured by the cheaper process may be. So he
randomly selected 16 picture tubes and their life time in
1 Select Stat 5 Click Options hours is given below.
2 Select Basic Statistics 6. Enter Confidence Level
3115 3250 3300 3156 3225 3400 3100 3118
3 Select 1-Sample t 7. Click OK
3095 3425 3256 3308 3115 3348 3125 3220
4 Select Data Range
 Sample mean = 3222.25
Sample variance = 110.362 MICROSOFT EXCEL OUTPUT
t19,0.05 = 1.753 n= 16
Column1

"
• Lower Limit =  − . #$

Mean 3222.25

Standard Error 27.59023076
= 3225.25 - 1.753* 110.36/  Median 3222.5
Mode 3115
= 3173.883 Standard Deviation 110.360923
Sample Variance 12179.53333
"
• Upper Limit =  + . #$



Kurtosis -1.017054845
Skewness 0.476199807
= 3225.25 + 1.753* 110.36/  Range 330
Minimum 3095
= 3270.367 Maximum 3425
Sum 51556
95% confidence interval for µ = [3173.88, 3270.37] Count 16
Confidence Level(90.0%) 48.367063

SHARE PRICE EXAMPLE

MINITAB OUTPUT
An investor wants to get an idea about the fluctuation of
One-Sample T: Life Time daily share prices of leading commercial banks in Sri
Variable N Mean StDev SE Mean 90% CI Lanka. He obtained the daily share prices of one
particular bank over a period of 25 days and is given
Life Time 16 3222.25 110.36 27.59 (3173.88, 3270.62) below.

127 125 130 132 135 121 129 136 132

135 142 138 121 126 134 137 119 126
123 132 140 129 133 137 128

Obtain 99% confidence interval for the mean daily share

price of this bank. Comment on your findings.
 Sample mean = 130.68
Sample variance = 38.3932 CONFIDENCE INTERVAL FOR P
t19,0.05 = 2.797 n= 25

"
• Lower Limit =  − . ##




= 130.68 - 2.797* 38.393/ 

= 127.21
"
• Upper Limit =  + . ##




= 130.68 + 2.797* 38.393/ 

= 134.15
95% confidence interval for µ = [127.21, 134.15]