Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals...

13
Lecture 1 Interval Estimation Dr. Hoda Ragab Rezk

Transcript of Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals...

Page 1: Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals (1)Pivotal Quantity Method (2)Maximum Likelihood Estimator (MLE) Method (3)Bayesian Method

Lecture 1

Interval Estimation

Dr. Hoda Ragab Rezk

Page 2: Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals (1)Pivotal Quantity Method (2)Maximum Likelihood Estimator (MLE) Method (3)Bayesian Method

2

The interval estimator of θ is called

a 100(1 − 𝛼)% confidence interval for θ if

P(L ≤ θ ≤ U) = 1 − 𝛼.

The L is called the lower confidence limit

The U is called the upper confidence limit.

The number (1− 𝛼) is called the confidence

coefficient or degree of confidence.

Page 3: Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals (1)Pivotal Quantity Method (2)Maximum Likelihood Estimator (MLE) Method (3)Bayesian Method

3

Remark The length of confidence interval =

upper confidence limit - lower confidence limit

Page 4: Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals (1)Pivotal Quantity Method (2)Maximum Likelihood Estimator (MLE) Method (3)Bayesian Method

Methods for constructing

confidence intervals

(1)Pivotal Quantity Method

(2)Maximum Likelihood Estimator (MLE) Method (3)Bayesian Method

(4)Invariant Methods

(5)Inversion of Test Statistic Method

(6)The Statistical or General Method.

Page 5: Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals (1)Pivotal Quantity Method (2)Maximum Likelihood Estimator (MLE) Method (3)Bayesian Method

Pivotal Quantity Method

Definition

Let X1, X2, ..., Xn be a random sample of size n

from a population X with probability density function

f(x; θ),

where θ is an unknown parameter.

A pivotal quantity Q is a function of X1,X2, ...,Xn and θ

whose probability distribution is independent of the

parameter θ.

5

Page 6: Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals (1)Pivotal Quantity Method (2)Maximum Likelihood Estimator (MLE) Method (3)Bayesian Method

Remark 1 The pivotal quantity Q(X1,X2, ...,Xn, θ) will

usually contain both the parameter θ and an

estimator (a statistic) of θ.

6

Page 7: Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals (1)Pivotal Quantity Method (2)Maximum Likelihood Estimator (MLE) Method (3)Bayesian Method

Example 1

Let X1, X2, ..., X11 be a random sample of size 11

from a normal distribution with unknown mean μ and

variance 𝜎2 = 9.9. If 𝑥𝑖11𝑖=1 =132 , then what is the

95% confidence interval for μ?

7

Page 8: Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals (1)Pivotal Quantity Method (2)Maximum Likelihood Estimator (MLE) Method (3)Bayesian Method

8

Answer 1

Since 𝑋𝑖 ∼ 𝑁(𝜇, 9.9), the confidence interval for 𝜇 is given by

X − zα2

σ

n, X + zα

2

σ

n

Since 𝑥𝑖11𝑖=1 = 132, the sample mean 𝑥 =

132

11= 12,

𝜎2

𝑛 =

9.9

11= 0.9

Since 1 − 𝛼 = 0.95 , 𝛼 = 0.05, So 𝑧𝛼2= 𝑧0.025 = 1.96

(from normal table)

Page 9: Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals (1)Pivotal Quantity Method (2)Maximum Likelihood Estimator (MLE) Method (3)Bayesian Method

9

The confidence interval for μ is

12 − 1.96 0.9 , 12 + 1.96 0.9

10.141, 13.859

Example 2

Let X1, X2, ..., X11 be a random sample of size 11

from a normal distribution with unknown mean μ and

variance 𝜎2 = 9.9. If 𝑥𝑖11𝑖=1 =132, then for what value of the

constant k is 12 − 𝑘 0.9 , 12 + 𝑘 0.9

a 90% confidence interval for μ?

Page 10: Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals (1)Pivotal Quantity Method (2)Maximum Likelihood Estimator (MLE) Method (3)Bayesian Method

Answer 2

The 90% confidence interval for μ when the variance

is given is

X − zα2

σ

n, X + zα

2

σ

n

Thus we need to find X ,σ

n and zα

2 corresponding to

1 − 𝛼 = 0.90. Hence

𝑥 =132

11= 12

𝜎2

𝑛 =

9.9

11= 0.9

10

Page 11: Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals (1)Pivotal Quantity Method (2)Maximum Likelihood Estimator (MLE) Method (3)Bayesian Method

𝑧𝛼2= 𝑧0.05 = 1.64 (from normal table).

Hence, the confidence interval for μ at 90%

confidence level is

12 − 1.64 0.9 , 12 + 1.64 0.9

10.444, 13.556

Comparing this interval with the given interval, we get

11

Page 12: Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals (1)Pivotal Quantity Method (2)Maximum Likelihood Estimator (MLE) Method (3)Bayesian Method

K=1.64

and the corresponding 90% confidence interval is

[10.141, 13.859]

Remark

• The confidence level is directly proportional to the length

of the confidence interval.

For example

• Notice that the length of the 90% confidence interval for μ

is 3.112. However, the length of the 95% confidence

interval is 3.718

12

Page 13: Lecture 1 Interval Estimation - fsalamri · Methods for constructing confidence intervals (1)Pivotal Quantity Method (2)Maximum Likelihood Estimator (MLE) Method (3)Bayesian Method

13

Thank You