[go: up one dir, main page]
Scribd
Upload
13 views9 pages

L2 Sampling Distribution

The document provides an overview of three key probability distributions related to functions of independent standard normal random variables: Chi-Square, t-Distribution, and F-Distribution. It details the definitions, properties, and applications of these distributions, including their probability density functions and moments. The content is aimed at students of Statistical Methods and Time Series Analysis, highlighting their significance in hypothesis testing across various scientific disciplines.

Uploaded by

Gaurab Das
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
13 views9 pages

L2 Sampling Distribution

The document provides an overview of three key probability distributions related to functions of independent standard normal random variables: Chi-Square, t-Distribution, and F-Distribution. It details the definitions, properties, and applications of these distributions, including their probability density functions and moments. The content is aimed at students of Statistical Methods and Time Series Analysis, highlighting their significance in hypothesis testing across various scientific disciplines.

Uploaded by

Gaurab Das
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 9

Review of Sampling Distributions

Course: Statistical Methods and Time Series Analysis (MA593)

Palash Ghosh

Department of Mathematics, IITG

Odd-Semester, Jul-Nov 2025

1/9
Sampling Distributions

• Here we will discuss three probability distributions which are the


distributions of some functions of independent standard normal
random variables.
• Chi-Square Distribution
• t-Distribution
• F-Distribution

2/9
Chi-Square Distribution

• Let 𝑋 ∼ 𝑁 (0, 1). What is the distribution of 𝑍 = 𝑋 2 ?


• Show that the random variable 𝑍 has the density function:
𝑧
𝑒− 2 𝑧 (1/2)−1
𝑓𝑍 (𝑧) = ,0 < 𝑧 < ∞
21/2 Γ(1/2)
= 0 otherwise

• The above probability density function is called the Chi-square


distribution with one degrees of freedom and it is denoted by 𝜒21 .
• Thus, if 𝑋 has standard normal distribution, then 𝑋 2 has the
Chi-square distribution with one degrees of freedom.

3/9
Chi-Square Distribution

• Show that the MGF of 𝑍 = 𝑋 2 is


1
𝜓𝑍 (𝑡) = (1 − 2𝑡)(−1/2) , for 𝑡 <
2

• Let 𝑋𝑖 , 𝑖 = 1, ⋯ , 𝑛 are independent standard normal random


variables. Using the MGF show that 𝑍𝑛 = 𝑋12 + ⋯ + 𝑋𝑛2 has the
Chi-square distribution with 𝑛 degrees of freedom.
• Show that 𝐸(𝑍𝑛 ) = 𝑛 and 𝑉 (𝑍𝑛 ) = 2𝑛.
• If 𝑍1 ∼ 𝜒2𝑚 , 𝑍2 ∼ 𝜒2𝑛 and they are independent, then
𝑍1 + 𝑍2 ∼ 𝜒2𝑚+𝑛 .

4/9
Chi-Square Distribution

• Let 𝑌𝑖 , 𝑖 = 1, ⋯ , 𝑛 are iid sample from 𝑁 (𝜇, 𝜎2 ) with 𝑛 > 1.


Define

̄ 1 𝑛 1 𝑛
2
𝑌 = ∑ 𝑌𝑖 , and 𝑆 = ∑(𝑌 − 𝑌 ̄ )2 .
𝑛 𝑖=1 𝑛 − 1 𝑖=1 𝑖

Here 𝑌 ̄ is the sample mean and 𝑆 2 is the sample variance. Then


(i) 𝑌 ̄ and 𝑆 2 are independent random variables.
2
(ii) 𝑌 ̄ ∼ 𝑁 (𝜇, 𝜎𝑛 ) and
(𝑛−1)𝑆 2
(iii) 𝜎2 ∼ 𝜒2𝑛−1 .
• Use: in Chi-square test. (Widely used across all disciplines of
science)

5/9
t-Distribution
• Let 𝑋 ∼ 𝑁 (0, 1) and 𝑌 ∼ 𝜒2𝑛 . Assume that 𝑋 and 𝑌 are
independent. Define
𝑋
𝑇 = .
√𝑌 /𝑛

• The distribution of the random variable 𝑇 is called the


𝑡-distribution with 𝑛 degrees of freedom.
• The above distribution is also known as Student’s 𝑡-distribution
• The probability density function of 𝑇 is given by

Γ( 𝑛+1
2 ) − 𝑛+1
𝑡2 2

𝑓(𝑡) = √ (1 + ) , −∞ < 𝑡 < ∞.


𝑛𝜋Γ( 𝑛2 ) 𝑛

6/9
t-Distribution

• The pdf of 𝑡-distribution is symmetric about zero and for 𝑛 > 1,


mean = median = mode = 0.
• MGF does not exist for t-distribution.
• Use: 𝑡-distribution is used for comparing means of two normal
distributions in problems of testing of hypotheses.

7/9
F-Distribution
• Let 𝑋 and 𝑌 are two independent random variables such that
𝑋 ∼ 𝜒2𝑚 and 𝑌 ∼ 𝜒2𝑛 .
• Define
𝑋/𝑚
𝑊 = .
𝑌 /𝑛

• The distribution of the random variable 𝑊 is called 𝐹 -distribution


with 𝑚 and 𝑛 degrees of freedom.
• The pdf of the 𝐹 -distribution is given by

Γ( 𝑚+𝑛
2 )𝑚
𝑚/2 𝑛/2
𝑛 𝑤(𝑚/2)−1
𝑓(𝑤) = 𝑚+𝑛 , if 𝑤 > 0
Γ( 𝑚 𝑛
2 )Γ( 2 ) (𝑚𝑤 + 𝑛) 2
= 0 otherwise

8/9
F-Distribution

• 𝐹 -distribution with 𝑚 and 𝑛 degrees of freedom is different


from 𝐹 -distribution with 𝑛 and 𝑚 degrees of freedom.
• Use:
a) 𝐹 -distribution plays a major role in statistical inference. It is used
in comparing the means of several normal populations in problems
of testing of hypotheses.
b) It is also used in comparing variances of normal populations in
problems of testing of hypotheses.

9/9

You might also like