ESTIMATION OF PARAMETERS:

In this chapter, we discuss the parameters which are associated with the probability distribution of
random variable X.
Definition: Let X be a random variable with some probability distribution depending on an
unknown parameter θ. Let 𝑋1 , 𝑋2 , 𝑋3 , … 𝑋𝑛 be sample of size 𝑛 taken from distribution of X. If
𝑔(𝑋1 , 𝑋2 , 𝑋3 , … 𝑋𝑛 ) is a function of sample to be used for estimating θ. We refer 𝑔 as an estimator
of θ. The value of 𝑔 assumes will be refer as an estimate of θ. We write 𝜃̂ = 𝑔(𝑋1 , 𝑋2 , 𝑋3 , … 𝑋𝑛 ).
Definition: Let 𝜃̂ be an estimate for the unknown parameter θ associated with the distribution of
random variable X. Then 𝜃̂ is an unbiased estimator for θ if 𝐸(𝜃̂) = θ , ∀ θ .

Note: Any good estimate should be close to the value it is estimating, unbiasedness refers the
average value of the estimate will be close to the true parameter value.

Definition: Let 𝜃̂ be an estimate of the parameter θ we say that 𝜃̂ is a consistent estimate of θ if

lim 𝑃{|𝜃̂ − θ| > 𝜖} = 0 ∀𝜖 > 0 𝑜𝑟 lim 𝑃{|𝜃̂ − θ| ≤ 𝜖} = 1 .
𝑛→∞ 𝑛→∞

Note: As sample size increases the estimate becomes ‘better’ is indicated in above definition.
We shall find unbiasedness and consistent of estimate using the following theorem.

Theorem: Let 𝜃̂ be an estimate of the parameter θ based on a sample size 𝑛. If 𝐸(𝜃̂) = θ,

lim 𝑉(𝜃̂) = 0 then 𝜃̂ is a consistent estimate of θ.
𝑛→∞
Proof: We shall prove by using Chebyshev’s inequality.
1 2 1
∴ lim 𝑃{|𝜃̂ − θ| > 𝜖} ≤ 2 𝐸(𝜃̂ − θ) = 2 𝐸{(𝜃̂ − E(𝜃̂)) + (E(𝜃̂) − θ)}2(Add &subtract E(𝜃̂))
𝑛→∞ 𝜖 𝜖
1 2
= 𝜖2 {𝐸[𝜃̂ − E(𝜃̂)] + 2𝐸{[𝜃̂ − E(𝜃̂) ](E(𝜃̂) − θ)} + E(E(𝜃̂) − θ)2 }
1 2
=𝜖2 {𝑉(𝜃̂) + 2𝐸(𝜃̂ − E(𝜃̂)(E(𝜃̂) − θ) + (E(𝜃̂) − θ) }
1 2
= 𝜖2 {𝑉(𝜃̂) + (E(𝜃̂) − θ) } → 0 𝑎𝑠 𝑛 → ∞ using given condition.
∴ 𝜃̂ is a consistent estimate of θ.
Examples:

1. Show that sample mean is an unbiased and consistent estimate of population mean.

Solution: Let 𝑋1 , 𝑋2 , … 𝑋𝑛 are the samples taken from the distribution of X having mean
value 𝜇.

∑𝑛
𝑖=1 𝑋𝑖
∴ 𝑋̅ = is the sample mean.
𝑛
To prove: 𝐸 (𝑋̅ ) = 𝜇 𝑎𝑛𝑑 𝑉(𝑋̅ ) = 0.

∑ 𝑛 ∑𝑛
𝑋𝑖 1 𝑖=1 𝑋𝑖
Consider 𝐸 (𝑋̅ ) = 𝐸 ( 𝑖=1 )= . ∑𝑛𝑖=1 𝑋𝑖 = = 𝜇.
𝑛 𝑛 𝑛

That is, the sample mean is an unbiased estimate of population mean.

𝑛 𝑛
∑𝑛𝑖=1 𝑋𝑖 1 1 𝑛𝑉(𝑋) 𝜎 2
𝑉(𝑋̅) = 𝑉 ( ) = 2 𝑉 (∑ 𝑋𝑖 ) = 2 ∑ 𝑉(𝑋𝑖 ) = = .
𝑛 𝑛 𝑛 𝑛2 𝑛
𝑖=1 𝑖=1

𝜎2
∴ lim 𝑉(𝑋̅) = lim =0
𝑛→∞ 𝑛→∞ 𝑛

That is, the sample mean is consistent estimate of population mean.

2. Show that sample variance 𝑆 2 is not an unbiased estimate of population variance.

Solution: Let 𝜎 2 be the variance for the distribution of X. That is, population variance. Let
𝑆 2 be the sample variance.

To prove: 𝐸(𝑆 2 ) ≠ 𝜎 2

∑𝑛 ̅ 2
𝑖=1(𝑋𝑖 −𝑋) ∑𝑛 2 ̅ 2 ̅
𝑖=1(𝑋𝑖 +(𝑋 ) −2𝑋 𝑋𝑖 ) [ ∑𝑛 2 ̅ 2 ̅2
𝑖=1(𝑋𝑖 ) ] +𝑛(𝑋 ) −2𝑛(𝑋 )
By definition, 𝑆 2 = = =
𝑛 𝑛 𝑛

∑𝑛 2
𝑖=1(𝑋𝑖 )
= − (𝑋̅)2 { 𝑠𝑖𝑛𝑐𝑒 ∑𝑛𝑖=1(𝑋̅)2 = 𝑛(𝑋̅)2 𝑎𝑛𝑑 𝑋̅ 𝑛 = ∑𝑛𝑖=1 𝑋𝑖 }
𝑛

∑𝑛 2
𝑖=1(𝑋𝑖 ) ∑𝑛 2
𝑖=1(𝑋𝑖 )
Consider, E(𝑆 2 ) = 𝐸 { − ̅̅̅
(𝑋)2 } = 𝐸 { ̅̅̅ )2 }
} − 𝐸{(𝑋
𝑛 𝑛

1
̅̅̅)2 }
= 𝑛 𝐸(∑𝑛𝑖=1(𝑋𝑖 2 )) − {𝐸(𝑋 [ 𝑠𝑖𝑛𝑐𝑒 𝐸(𝑋 2 ) = 𝑉(𝑋) + 𝜇 2 , 𝐸(𝑋̅ 2 ) = 𝑉(𝑋̅) + 𝜇 2 ]

1 𝜎2 𝜎2
= 𝑛 [𝑛 (𝜎 2 + 𝜇 2 )] − { 𝑛 + 𝜇 2 } (sample variance is )
𝑛

𝜎2 𝜎2 𝑛−1
= (𝜎 2 + 𝜇 2 ) − { 𝑛 + 𝜇 2 } = (𝜎 2 ) − { 𝑛 } = 𝜎 2 ( )
𝑛

∴ 𝐸(𝑆 2 ) ≠ 𝜎 2
 𝑥
1 −
3. Show that 𝑋̅ is a random sample of size 𝑛, 𝑓(𝑥, 𝜃) = {𝜃 𝑒 0 < 𝑥 < ∞, 0 < 𝜃 < ∞
𝜃

0 𝑒𝑙𝑠𝑒𝑤ℎ𝑒𝑟𝑒
𝜃2
is an unbiased estimate of 𝜃 and has variance .
𝑛

Solution: First we shall find the mean and variance of X.

∞ ∞ 𝑥
1
𝐸(𝑋) = ∫−∞ 𝑥 𝑓(𝑥, 𝜃) 𝑑𝑥 = ∫0 𝑥 𝑒 − 𝜃 𝑑𝑥
𝜃
𝑥 𝑥
1 ∞ −𝑥 1 𝑒− 𝜃 𝑒− 𝜃 ∞ 𝜃2
= ∫ 𝑥 𝑒 𝜃 𝑑𝑥 = { 𝑥 − }0 = =𝜃
𝜃 0 𝜃 −1 1 𝜃
𝜃 𝜃2

∞ ∞
1 −𝑥
𝐸(𝑋 2 ) = ∫ 𝑥 2 𝑓(𝑥, 𝜃) 𝑑𝑥 = ∫ 𝑥 2 𝑒 𝜃 𝑑𝑥
−∞ 0 𝜃
𝑥 𝑥 𝑥
1 ∞ 2 −𝑥 1 2
𝑒− 𝜃 𝑒− 𝜃 𝑒− 𝜃 ∞ 2𝜃 3
= ∫ 𝑥 𝑒 𝜃 𝑑𝑥 = { 𝑥 − 2𝑥 + 2 }0 = = 2𝜃 2
𝜃 0 𝜃 −1 1 −1 𝜃
𝜃 𝜃2 𝜃3
∴ 𝑉(𝑋) = 2𝜃 2 − 𝜃 2 = 𝜃 2
𝜃2
The random variable 𝑋 has mean 𝜃 and variance 𝜃 2 . Hence, (𝑋̅) = 𝜃 , 𝑉(𝑋̅) = 𝑛

4. Let 𝑋1 , 𝑋2 , … 𝑋𝑛 are the samples taken from a normal distribution with 𝜇 = 0 and
∑ 𝑥𝑖 2
variance 𝜎 2 = 𝜃, 0 < 𝜃 < ∞. Show that 𝑌 = is an unbiased and consistent
𝑛

estimate of 𝜃.

𝜃
Solution: Let 𝑋 has 𝑁( 0, 𝜃) and sample mean (𝑋̅) has 𝑁( 0, )
𝑛

𝜃
Therefore, (𝑋̅) = 0 , 𝑉(𝑋̅) = 𝑛
∑ 𝑥𝑖 2
Let 𝑌 = .
𝑛

To prove 𝐸(𝑌) = 0 and 𝑉(𝑌) = 0 𝑎𝑠 𝑛 → ∞

∑ 𝑋𝑖 2 1 𝑛 𝐸(𝑋 2 )
Consider 𝐸(𝑌) = 𝐸 ( ) = 𝐸(∑ 𝑋𝑖 2 ) = = 𝐸(𝑋 2 )
𝑛 𝑛 𝑛

𝐸(𝑌) = 𝑉(𝑋) + [ 𝐸(𝑋)]2 = 𝜃 which implies Y has unbiased estimate of 𝜃.

 𝑋− 𝜇 𝑋− 𝜇 2 𝑋− 0 2 𝑋2
We know that, X has ( 𝜇, 𝜎 2 ) , 𝑍 = has 𝑁( 0, 1) and 𝑍 2 = ( ) =( ) =
𝜎 𝜎 √𝜃 𝜃

has 𝜒 2 (1)
𝑋2 𝑋2
∴ 𝐸 ( ) = 1 𝑎𝑛𝑑 𝑉 ( ) = 2
𝜃 𝜃
This implies 𝐸(𝑋 2 ) = 𝜃 , 𝑉(𝑋 2 ) = 2𝜃 2
∑ 𝑋𝑖 2 1 𝑛 𝑉(𝑋 2 ) 𝑉(𝑋 2 ) 2𝜃2
Consider 𝑉(𝑌) = 𝑉 ( ) = 𝑛2 , 𝑉(∑ 𝑋𝑖 2 ) = = =
𝑛 𝑛2 𝑛 𝑛
2𝜃2
∴ lim 𝑉(𝑌) = lim = 0
𝑛→∞ 𝑛→∞ 𝑛

5. Let 𝑌1 , 𝑌2 be two independent unbiased statistics for 𝜃. The variance of 𝑌1 is twice the
variance of 𝑌2 . Find the constants 𝑘1 𝑎𝑛𝑑 𝑘2 such that 𝑌 = 𝑘1 𝑌1 + 𝑘2 𝑌2 is an unbiased
statistics for 𝜃 with smallest possible variance for such a linear combination.
Solution: Given that 𝐸(𝑌1 ) = 𝐸(𝑌2 ) = 𝜃, 𝑉(𝑌1 ) = 2𝑉( 𝑌2 ) = 2𝜎 2 .
To find 𝑘1 𝑎𝑛𝑑 𝑘2 such that 𝐸(𝑌) = 𝐸(𝑘1 𝑌1 + 𝑘2 𝑌2 ) = 𝜃.
That is, 𝑘1 𝐸(𝑌1 ) + 𝑘2 𝐸(𝑌2 ) = 𝜃 ⇒ 𝑘1 𝜃 + 𝑘2 𝜃 = 𝜃 ⇒ 𝑘1 + 𝑘2 = 1 ⇒ 𝑘2 = 1 − 𝑘1 .
2 2
Consider 𝑉(𝑌) = 𝑉(𝑘1 𝑌1 + 𝑘2 𝑌2 ) = 𝑘1 2 𝑉(𝑌1 ) + 𝑘2 𝑉(𝑌2 ) = 𝜎 2 (2𝑘1 2 + 𝑘2 )
2
𝑉(𝑌) = 𝜎 2 (2𝑘1 2 + 𝑘2 ) = 𝜎 2 (2𝑘1 2 + (1 − 𝑘1 )2 )
𝑑𝑉(𝑌)
V(Y) has minima if =0
𝑑𝑘1

𝑑𝑉(𝑌) 𝑑[𝜎2 (2𝑘1 2 +(1−𝑘1 )2 )] 1 2

⇒ = = 0 ⇒ 4𝑘1 − 2(1 − 𝑘1 ) = 0 ⇒ 𝑘1 = 3 and 𝑘2 = 3
𝑑𝑘1 𝑑𝑘1

6. Let 𝑋1 , 𝑋2 . . 𝑋25 , 𝑌1 , 𝑌2 … 𝑌25 be two independent random samples from the

𝑋̅
distribution 𝑁(3, 16) , 𝑁(4, 9) respectively. Evaluate P ( 𝑌̅ > 1)

16 9
Solution: Let 𝑋 ~𝑁(3, 16), 𝑌~𝑁(4, 9). Then, 𝑋̅ ~𝑁 (3, ) , 𝑌̅ ~𝑁(4, )
25 25

𝑋̅ 16
Now 𝑌̅ > 1 ⇒ 𝑋̅ > 𝑌̅ ⇒ 𝑋̅ − 𝑌̅ > 0. Since, 𝑋̅ − 𝑌̅ ~ 𝑁[ 3 × 1 + 4 × (−1), 12 × 25 +
9
(−1)2 × ]~𝑁(−1, 1) and Z= 𝑋̅ − 𝑌̅ + 1 ~𝑁(0, 1)
25

𝑋̅
Consider P (𝑌̅ > 1) = 𝑃(𝑋̅ − 𝑌̅ > 0) = 𝑃(𝑋̅ − 𝑌̅ + 1 > 1) = 𝑃(𝑍 > 1)

= 1 − 𝛷(1)=1- 0.841=0.159
Interval estimation:

Let ‘X’ be a random variable with some probability distribution, depending on an unknown
parameter θ. An estimate of θ given by two magnitudes within which θ can lie is called an interval
estimate of the parameter θ. The process of obtaining an interval estimate for θ is called interval
estimation.

Let θ be an unknown parameter to be determined by a random sample 𝑋1 , 𝑋2 , 𝑋3 , … 𝑋𝑛 of size 𝑛.

The confidence interval for the parameter θ is a random interval containing the parameter with
high probability say 1 − 𝛼; 1 − 𝛼 is called confidence coefficient.

Suppose that 𝑃{(𝐻(𝑋1 , 𝑋2 , … 𝑋𝑛 ) < 𝜃 < 𝐺(𝑋1 , 𝑋2 , … 𝑋𝑛 )} = 1 − 𝛼 then

{(𝐻(𝑋1 , 𝑋2 , … 𝑋𝑛 ), 𝐺(𝑋1 , 𝑋2 , … 𝑋𝑛 )} is a [(1 − 𝛼) × 100]% confidence interval.

Note: Let 𝑋1 , 𝑋2 . . 𝑋𝑛 be a random sample of size 𝑛 from a normal distribution 𝑁(𝜇, 𝜎 2 ).

∑𝑛
𝑖=1 𝑋𝑖 𝜎2
1. 𝑋̅ = ~ 𝑁(𝜇, ) for 𝜇, 𝜎 2 is known.
𝑛 𝑛
𝑋̅ −𝜇
2. 𝑇 = 𝑆 ~ 𝑇(𝑛 − 1) for 𝜇, 𝜎 2 is unknown.
√𝑛−1

∑𝑛 ̅ 2
𝑖=1(𝑋𝑖 −𝑋)
3. 𝑌 = ~𝜒 2 (n) for 𝜎 , 𝜇 is known
𝜎2
𝑛𝑆 2
4. 𝑌 = ~𝜒 2 (n-1) for 𝜎 , 𝜇 is unknown
𝜎2

Confidence Interval for mean:

∑𝑛
𝑖=1 𝑋𝑖 𝜎2
1) 𝜎 2 is known: Consider 𝑋̅ = ~ 𝑁(𝜇, )
𝑛 𝑛

𝑋̅ − 𝜇
∴𝑍= 𝜎 ~𝑁(0,1)
√𝑛
To find 𝑎 such that 𝑃(−𝑎 < 𝑍 < 𝑎) = 1 − 𝛼

𝑋̅ − 𝜇
⇒ 𝑃 (−𝑎 < 𝜎 < 𝑎) = 1 − 𝛼
√𝑛
𝑎𝜎 𝑎𝜎
⇒ 𝑃 (𝑋̅ − < 𝜇 < 𝑋̅ + ) = 1 − 𝛼
√𝑛 √𝑛
𝑎𝜎 𝑎𝜎
⇒ 𝜇 ∈ (𝑋̅ − , 𝑋̅ + )
√𝑛 √𝑛

Examples:
1. Let the observed value of 𝑋̅ of size 20 from a normal distribution with 𝜇 and 𝜎 2 =80 be
81.2. Obtain 95% confidence interval for the mean 𝜇.

80
Solution: Let 𝑋~ 𝑁(𝜇, 80) , 𝑋̅ ~ 𝑁 (𝜇, 20) = 𝑁(𝜇, 4)

𝑋̅ − 𝜇 𝑋̅ − 𝜇
∴𝑍= 𝜎 = ~𝑁(0,1)
2
√𝑛
1.95
𝑃(−𝑎 < 𝑍 < 𝑎) = 0.95 ⇒ 2 𝛷(𝑎)-1=0.95 ⇒ 𝛷(𝑎)= = 0.975 ⇒ 𝑎 = 1.96
2

𝑎𝜎 𝑎𝜎
⇒ 𝜇 ∈ (𝑋̅ − , 𝑋̅ + ) ∈ (81.2 − 1.96 × 2, 81.2 + 1.96 × 2)
√𝑛 √𝑛

⇒ 𝜇 ∈ (77.28, 85.12)

2. Let the observed value of 𝑋̅ of size 𝑛 from a normal distribution with 𝜇 and 𝜎 2 = 9. Find 𝑛
such that 𝑃(𝑋̅ − 1 < 𝜇 < 𝑋̅ + 1) = 0.9 approximately.

9
Solution: Let 𝑋~ 𝑁(𝜇, 9) , 𝑋̅ ~ 𝑁 (𝜇, 𝑛)

𝑋̅ − 𝜇
∴𝑍= ~𝑁(0,1)
3
√𝑛
1.9
∴ 𝑃(−𝑎 < 𝑍 < 𝑎) = 0.9 ⇒ 2 𝛷(𝑎)-1=0.9 ⇒ 𝛷(𝑎)= 2 = 0.95 ⇒ 𝑎 = 1.65

Given that, 𝑃(𝑋̅ − 1 < 𝜇 < 𝑋̅ + 1) = 0.9 ⇒ 𝑃(1 − 𝑋̅ > −𝜇 > −𝑋̅ − 1) = 0.9
⇒ 𝑃(1 > 𝑋̅ − 𝜇 > −1) = 0.9
⇒ 𝑃(−1 < 𝑋̅ − 𝜇 < 1) = 0.9
−1 𝑋̅ − 𝜇 1
⇒ 𝑃( < < ) = 0.9
3⁄√𝑛 3⁄√𝑛 3⁄√𝑛
−1 1
⇒ 𝑃( <𝑍< ) = 0.9
3⁄√𝑛 3⁄√𝑛
√𝑛
⇒ 2𝛷 ( ) − 1 = 0.9
3
 √𝑛
⇒ 𝛷( ) = 1.9⁄2 = 0.95
3
√𝑛
⇒( ) = 1.65 ⇒ √𝑛 = 4.95 ⇒ 𝑛 = 24.5025 ≅ 25
3
2) 𝜎 2 is unknown:

𝑋̅ −𝜇
Consider 𝑇 = 𝑆 ~𝑇(𝑛 − 1) is in t- distribution with (n-1) degrees of freedom.
√𝑛−1

To find 𝑎 such that 𝑃(−𝑎 < 𝑇 < 𝑎) = 1 − 𝛼

𝑋̅ − 𝜇
⇒ 𝑃 (−𝑎 < < 𝑎) = 1 − 𝛼
𝑆
√𝑛 − 1
𝑎𝑆 𝑎𝑆
⇒ 𝑃 (𝑋̅ − < 𝜇 < 𝑋̅ + )= 1−𝛼
√𝑛 − 1 √𝑛 − 1
𝑎𝑆 𝑎𝑆
⇒ 𝜇 ∈ (𝑋̅ − , 𝑋̅ + )
√𝑛 − 1 √𝑛 − 1

Examples:

1. Let a random sample of size 17 from 𝑁(𝜇, 𝜎 2 ) yields 𝑋̅=4.7 and 𝑆 2 =5.76. Determine
90% confidence interval for 𝜇.

Solution: Given that 𝑛 = 17, 𝑋̅=4.7 and 𝑆 2 =5.76

𝑋̅ −𝜇 4(𝑋̅ −𝜇)
Let 𝑇 = 𝑆 = ~𝑇(17 − 1) ~𝑇(16)
√5.76
√𝑛−1

To find 𝑎 such that 𝑃(−𝑎 < 𝑇 < 𝑎) = 0.90 ⇒ 2𝛷(𝑎) − 1 = 0.90 ⇒ 𝑎 = 1.75

1.75 × √5.76 1.75 × √5.76

∴ 𝜇 ∈ (4.7 − , 4.7 + ) ⇒ 𝜇 ∈ (3.65, 5.75)
√16 √16

Confidence interval for variance:

1) 𝜇 is known:
∑𝑛 ̅ 2
𝑖=1(𝑋𝑖 −𝑋)
Let 𝑌 = ~𝜒 2 (n)
𝜎2
 To find a and b such that 𝑃(𝑎 < 𝑌 < 𝑏) = 1 − 𝛼
𝛼 𝛼
⇒𝑃(𝑌 < 𝑎) = , 𝑃(𝑌 > 𝑏) =
2 2
∑𝑛 ̅ 2
𝑖=1(𝑋𝑖 −𝑋)
∴ 𝑃 (𝑎 < < 𝑏) = 1 − 𝛼
𝜎2

1 𝜎2 1
⇒ 𝑃( < < )=1−𝛼
𝑏 𝑛 ̅
∑𝑖=1(𝑋𝑖 − 𝑋) 2 𝑎
∑𝑛𝑖=1(𝑋𝑖 − 𝑋̅)2 2
∑𝑛𝑖=1(𝑋𝑖 − 𝑋̅)2
⇒ 𝑃( < 𝜎 < )= 1−𝛼
𝑏 𝑎
∑𝑛 ̅ 2 ∑𝑛
𝑖=1(𝑋𝑖 −𝑋)
̅ 2
𝑖=1(𝑋𝑖 −𝑋) ∑ 𝑋𝑖
⇒ 𝜎2 ∈ ( , ) where 𝜇 = = 𝑋̅
𝑏 𝑎 𝑛

Examples: 1. If 8.6, 7.9, 8.3, 6.4, 8.4, 9.8, 7.2, 7.8, 7.5 are the observed values of a
random sample of size 9 from a distribution 𝑁(8, 𝜎 2 ), construct 90% confidence interval
for 𝜎 2 .
∑𝑛 ̅ 2
𝑖=1(𝑋𝑖 −𝑋 ) 1
Solution: Let 𝜇 = 8, 𝑛 = 9 and 𝑌 = = {(0.6)2 + (0.1)2 + (0.3)2 +
𝜎2 𝜎2
7.35
(1.6)2 + (0.4)2 + (1.8)2 + (0.8)2 + (0.2)2 + (0.5)2 } = ~𝜒 2 (9)
𝜎2

To find a and b such that 𝑃(𝑎 < 𝑌 < 𝑏) = 1 − 𝛼 = 0.90 ⇒ 𝛼 = 0.10

𝛼 0.10 𝛼 0.10
⇒𝑃(𝑌 < 𝑎) = = = 0.05 ⇒ 𝑎 = 3.33 , 𝑃(𝑌 > 𝑏) = = = 0.05
2 2 2 2

⇒ (𝑌 < 𝑏) = 1 − 0.05 = 0.95 ⇒ 𝑏 = 16.9 using chi square table for 9 degrees of
freedom.
∑𝑛 ̅ 2 ∑𝑛
𝑖=1(𝑋𝑖 −𝑋 )
̅ 2
𝑖=1(𝑋𝑖 −𝑋 ) 7.35 7.35
∴ 𝜎2 ∈ ( , ) = (16.9 , ) = (0.43, 2.21)
𝑏 𝑎 3.33

2) 𝜇 is unknown:
𝑛𝑆 2
Let 𝑌 = ~𝜒 2 (𝑛 − 1)
𝜎2

To find 𝑎 and 𝑏 such that 𝑃(𝑎 < 𝑌 < 𝑏) = 1 − 𝛼

𝛼 𝛼
⇒𝑃(𝑌 < 𝑎) = , 𝑃(𝑌 > 𝑏) =
2 2
𝑛𝑆 2
∴ 𝑃 (𝑎 < < 𝑏) = 1 − 𝛼
𝜎2

1 𝜎2 1
⇒ 𝑃( < 2
< )=1−𝛼
𝑏 𝑛𝑆 𝑎
𝑛𝑆 2 𝑛𝑆 2
⇒ 𝑃( < 𝜎2 < )=1−𝛼
𝑏 𝑎
 2
𝑛𝑆 2 𝑛𝑆 2
⇒ 𝜎 ∈( , )
𝑏 𝑎

Examples:

1. A random sample of size 15 from a normal distribution 𝑁(𝜇, 𝜎 2 ) yields 𝑋̅ =

3.2 , 𝑆 2 = 4.24 . Determine a 90% confidence interval for 𝜎 2 .

𝛼 𝑛𝑆 2
Solution: Given that,1 − 𝛼 = 0.9 ⇒ 𝛼 = 0.1 & = 0.05, 𝑌 = ~𝜒 2 (15 − 1) = 𝜒 2 (14)
2 𝜎2

𝛼 0.10 𝛼 0.10
∴ 𝑃(𝑌 < 𝑎) = = = 0.05 ⇒ 𝑎 = 6.57 , 𝑃(𝑌 > 𝑏) = = = 0.05
2 2 2 2
⇒ (𝑌 < 𝑏) = 1 − 0.05 = 0.95 ⇒ 𝑏 = 23.7 using chi square table for 14 degrees of
freedom.
15 × 4.24 15 × 4.24
∴ 𝜎2 ∈ ( , ) = (2.68, 9.68)
23.7 6.57
Extra Problems: 1. A random sample of size 9 from a normal distribution 𝑁(𝜇, 𝜎 2 )
yields 𝑆 2 = 7.63. Determine a 95% confidence interval for 𝜎 2 .
𝐴𝑁𝑆: 𝜎 2 ∈ (3.924, 31.5)
2. A random sample of size 15 from a normal distribution 𝑁(𝜇, 𝜎 2 ) yields 𝑋̅ =
3.2 , 𝑆 2 = 4.24 . Determine a 95% confidence interval for 𝜇.
𝐴𝑁𝑆: 𝜇 ∈ (2.02, 4.38)
3. A random sample of size 25 from a normal distribution 𝑁(𝜇, 4) yields 𝑋̅ =
78.3 , 𝑆 2 = 4.24 . Determine a 99% confidence interval for 𝜇.
𝐴𝑁𝑆: 𝜇 ∈ (77.268, 79.332)