Lecture Notes 6

Course No. STA 0542 3117A

Title: Statistics for Architects

Sampling Distribution

Prepared by
Professor Dr. Mohammad Romel Bhuia
Department of Statistics
SUST

1
Basic concepts
• Population
– the complete set of all objects or people of interest.
• Sample
– a set of data collected and/or selected from a
population by a defined procedure.
• Parameter
– the unknown values of an entire population, such as
the mean and standard deviation.
• Statistic
– any quantity computed from values in a sample that is
used for a statistical purpose.
2
Sampling Distribution
• Sampling Distribution
– Sampling distributions are probability
distributions of statistics.

• Example: Distribution of sample mean,

distribution of sample variance, etc.

3
Important Points
• If the original population is normally distributed, then
for any sample size n the distribution of the sample
mean is also normal. That is,

• If the distribution of the original population is not

known, but n is sufficiently “large”, the distribution
of the sample mean is approximately normal with
mean and variance given as

This result is known as the central limit theorem

(CLT).
4
 Chi-Square Distribution
Chi-Square Variate
• The square of a standard normal variate is
known as a chi-square variate with 1 degree of
freedom (d.f.). Thus,
X −
( )
If X ~ N  ,  , then Z =
2


~ N (0,1)

and 2
 X − 
Z =
2
 is a chi-square variate with 1 d.f.
  

5
Chi-Square Variate

• In general, if Xi, (i =1, 2, …, n) are n

independent normal variates with mean µi and
variance σi2 (i =1, 2, …, n), then
2
n
 X i − i 
 =  
2

i =1  i 
is a chi-square variate with n d.f.

6
Definition of Chi-Square Distribution

7
8
Properties of chi-square distribution

• It is a continuous distribution.
• The distribution has only one parameter, n.
• The shape of the distribution depends
upon the d.f, n.
• The mean of chi-square distribution with n
d.f. is n and the variance is 2n.
• If 𝑈 ∼𝜒2 with 𝑛 d.f. and 𝑉 ∼𝜒2 with m d.f.
are independent, then 𝑈 +𝑉 ∼ 𝜒2 with m+n
d.f.
9
Applications of chi-square distribution
• To test the significance of population
variance.
• To test the goodness of fit.
• To test the independence of attributes.
• To test the homogeneity of independent
estimates of the population variance.
• To test the homogeneity of independent
estimates of the population correlation
coefficient.
10
Student’s t-Distribution
Let xi (i=1, 2,…,n) be a random sample of size n from a
normal population with mean μ and variance σ2. Then
Student’s t is defined by the statistic
n

 ( xi − x )
2

x − 1 n
t= ; where, x =  xi , S 2 = i =1
s n i =1 n −1
n
It follows Student’s t-distribution with ν = (n-1) d.f.
with probability density function

1 1
f (t ) =  ; −  t  
1    +1
   ,   t2  2
 2 2  1 + 
  
11
Student’s t-Distribution
The probability density function of Student’s t-
distribution with ν = (n-1) d.f. is
1 1
f (t ) =  ; −  t  
1    +1
   ,   t2  2
 2 2  1 + 
  
If ν = 1, 1 1 1 1
f (t ) =  = 
 1 1  1+ t 2
 1+ t2
 , 
2 2
which is the p.d.f. of standard Cauchy distribution.
Hence, when ν = 1, Student’s t-distribution reduces to
Cauchy distribution.

12
Probability Curve: t-Distribution

13
Properties of t-distribution
• t-distribution ranges from -∞ to +∞, just
as in a normal distribution.

• The t-distribution is bell-shaped and

symmetric about 0.

• The mean of t-distribution is 0 if its d.f. is

greater than 1.

• t-distribution has a greater dispersion

than the standard normal distribution.

14
Application of t-distribution

• To test the significance of population mean.

• To test the significance in the difference
between two population means.
• To test the equality of two correlated means.
• To test the significance of population
correlation coefficient.
• To test the significance of the regression
coefficient.
• To test the equality of two independent
regression coefficients.

15
F-Distribution
F-distribution with n1 and n2 d.f. is the
distribution of the ratio of two independent
Chi-square variates divided by their
respective d.f. Thus the F-statistic may be
defined as
 n2
1
n1
F= where  n21 and  n22 are two independent
 n2
2
n2
 2 with n1 and n2 d.f., respectively.

16
F-Distribution
Hence the p. d. f. of F is given by
𝑛1
𝑛1 2 −1
1 𝐹
𝑛2
𝑓 𝐹 = 𝑛1 𝑛2 ⋅ 𝑛1 +𝑛2
𝛽 , 𝑛1 2
2 2 1+ 𝐹
𝑛2
𝑛1
𝑛1 2 𝑛1
−1
𝑛2 𝐹 2
= 𝑛1 𝑛2 ⋅ 𝑛1 +𝑛2 ,0 ≤ 𝐹 ≤ ∞
𝛽 , 𝑛1 2
2 2 1+ 𝐹
𝑛2

17
Curve: F-Distribution

18
Properties of F-distribution
• F-distribution ranges from 0 to +∞.

• The F-distribution is asymmetric.

𝑛2
• The mean of F-distribution is = , 𝑛2 > 2
𝑛2 −2
This is independent of 𝑛1 and is always greater than1.

• Mode of F-distribution is always less than 1.

• The moments of F distribution depend only on

𝑛1 and 𝑛2

19
Application of F-distribution

• To test the equality of two population

variances.
• To test the equality of several (more
than two) population means.
• In analysis of variance (ANOVA).
• In design and analysis of experiments.

20