Lecture NotConfidence Interval

CONFIDENCE INTERVAL
HIHGLIGHTS
• Importance
• Properties
• How to construct C.I
Introduction
•The purpose of statistics is to use information contained in a
sample drawn from a population to make estimate about
certain population parameter that characterizes the
population from the sample is drawn.
•Population parameters are measured values that can aid
information processing contained in the population. Examples
of Population parameter: Population mean, population
variance, population standard deviation and proportion.
We can call our population parameter of interest; the target
parameter
Estimator is a rule rule/formular that tells how to estimate
the values of a target population based on the on the
measurements contained in a sample data.
There are two methods of estimating/deriving estimator of a
target population parameter:
(i)Point estimator: a formula which determine a single value
for the target parameter
(ii)Interval estimator (Confidence interval): target parameter
lies between two values.
Point Estimator
•Many estimators (point or interval) can be obtained for a target
parameter through the different available methods (MLE, Least Square)
but some estimators are good and sum are bad while some are better.
•Choice of a good estimator must have the following properties
(i) Unbiasedness; where is the estimator and is the target population
parameter
(ii)Minimum variance from the population parameter
If the , then the estimator is said to be biased and the Bias .
If then is positively bias while then is negatively bias.
The average of the square of the distance between the estimator and its
corresponding target parameter: is called the mean square error.
It can be shown that

.
is unbiased estimator of the sample mean and is unbiased estimator of the sample
variance.
INTERVAL
•Properties of a good Interval Estimator
ESTIMATOR/
CONFIDENCE i. Contain the target population parameter
INTERVAL ii. It should be relatively narrow
•Our objective therefore, is to find an
Where is lower confidence limit and interval estimator capable of generating
is upper confidence limit. However, narrow interval that have high probability
different sample may produce of enclosing .
different interval estimator
•The probability that a randomly selected
confidence interval will enclose that target
parameter is called confidence coefficient.
Situation Sample Condition Standard deviation Confidence interval
statistics
Mean 𝜇 large 𝑛, 𝑥ҧ,𝑠 𝑛 > 30 𝑠 𝑠
𝑥ҧ
+ 𝑍𝛼 ൬ ൰
sample size ξ𝑛 2 ξ𝑛
Mean 𝜇 small 𝑛, 𝑥ҧ
,𝑠 𝑛 < 30 𝑠 𝑠
𝑥ҧ
+ 𝑡(𝛼 ,𝑛−1) ൬ ൰
sample size ξ𝑛 2 ξ𝑛
Proportion 𝑛, 𝑥ҧ
𝑜𝑟 𝑛, 𝑝 𝑛𝑝 > 10 or
𝑝ሺ1 − 𝑝ሻ 𝑝ሺ1 − 𝑝ሻ
𝑛(1 − 𝑝) > 30 ඨ + 𝑍𝛼 ቌ ඨ
𝑥ҧ ቍ
𝑛 2 𝑛
𝑠
Variance 𝜎 2 𝑛, 𝑠 2 or Population is or 𝑠 2 (𝑛 − 1) 𝑠 2 (𝑛 − 1)
nearly normal ξ 𝑛
2 , 2
𝜒 𝛼 𝜒 𝛼
𝑝ሺ1 − 𝑝ሻ (𝑛, ) (𝑛,1− )
2 2
ඨ
𝑛
Standard 𝑛, 𝑠 2 or 𝑠
deviation 𝜎 𝑠 2 (𝑛 − 1) 𝑠 2 (𝑛 − 1)
ξ𝑛 ඩ ,ඩ
𝜒2 𝛼 𝜒2 𝛼
(𝑛, ) (𝑛,1− )
2 2
Steps to construct confidence interval
• Determine sample size and Target parameter
• Determine the pivotal statistics
• Construct the confidence interval by to the

target parameter.
Types of confidence interval
i. , is two tail confidence interval

ii.,is lower one tail confidence interval
iii.,is lower one tail confidence interval
Sampling Distribution of the sample mean
/Distribution of sample mean
•Is the distribution of the means of the random sampling from
a population
•Uses of Distribution of sample mean
i. To compute probabilities
ii. To compute confidence interval for target population
parameter
iii.For hypothesis testing on the target population parameter
Central Limit Theorem (CLT)
•States that if all possible random samples, each of size n, are taken
from any population with mean and standard deviation , the sampling
distribution of the sample mean will
i. Have a mean
ii.
iii.Be normally distributed when the parent population is normally
distributed or will be approximately normally distributed for samples
of large sample size () when the parent population is not normally
distributed.
USES OF CLT
•To select a sample of size n from a finite population of size N, adjustment must be
made when is calculated in certain cases.
• Thus where is called finite population correction factor
•Meanwhile, if n is much smaller than N (), the will be very close to 1 and therefore
can be ignored when calculating and if n is , then, it is necessary to the correction
factor.
•For Example if and , we must use the correction factor.
Estimating population parameter using CLT
•Normal distribution
Student t distribution
•When the population variance is unknown and the distribution of the
underlying population is also unknown but is approximately normal, we
determine standard deviation from the sample data and use t
distribution

Lecture NotConfidence Interval

Uploaded by

Copyright:

Available Formats

Lecture NotConfidence Interval

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lecture NotConfidence Interval

Uploaded by

Copyright:

Available Formats

CONFIDENCE INTERVAL

It can be shown that

• Determine sample size and Target parameter

• Determine the pivotal statistics

• Construct the confidence interval by to the

i. , is two tail confidence interval

You might also like