Scribd
Upload
3 views1 page

STAT250 Lec6 2

This lecture discusses the application of the Central Limit Theorem to samples from a Bernoulli distribution, demonstrating that the normalized Binomial distribution converges to a standard Normal distribution as sample size increases. It provides examples related to defective items in a shipment and customer opinions, illustrating how to calculate probabilities and sample sizes for estimating proportions. Exercises are included for further practice on the concepts presented.

Uploaded by

ambre310506
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
3 views1 page

STAT250 Lec6 2

This lecture discusses the application of the Central Limit Theorem to samples from a Bernoulli distribution, demonstrating that the normalized Binomial distribution converges to a standard Normal distribution as sample size increases. It provides examples related to defective items in a shipment and customer opinions, illustrating how to calculate probabilities and sample sizes for estimating proportions. Exercises are included for further practice on the concepts presented.

Uploaded by

ambre310506
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

STAT 250 - Lecture 6 part 2

Samples from Bernoulli Experiment- (Sec 7.5)

If Y1 , . . . , Yn is a sample from Bernoulli(p) distribution, we can use Central Limit theorem to


conclude that Normalized Binomial distribution converges to standard Normal(0, 1).

In a Bernoulli experiment the trials are independent with same probability of success p and the
variable Yi records whether the ith trial is a success Yi = 1 or not Yi = 0. The variables Y1 , . . . Yn
form a sample from Bernoulli(p) distribution. If we sum them
S n = Y1 + · · · + Yn
we get the total number of success in n trials, which has the Binomial(n, p) distribution. If we
normalize it by subtracting the mean and dividing by the standard deviation
Sn − np
Zn = p
np(1 − p)
This is equivalent to
Sn /n − p Ȳ − p
Zn = p =p
p(1 − p)/n p(1 − p)/n
and since the mean of the Bernoulli(p) elements of the sample is E(Yi ) = p and its variance is
V (Yi ) = p(1 − p), the Central Limit Theorem shows
S − np Ȳ − p
p n =p ⇒ Normal(0, 1), as n → ∞.
np(1 − p) p(1 − p)/n
which can be used to give approximate probabilities for the Binomial distribution if n is large. The
rule for what is large enough for Bernoulli sample is: n > 9(max(p, 1 − p)/ min(p, 1 − p)).

7→ read Example 7.10, 7.11

Example 1 A machine usually produces about 10% defective items. A shipment of 100 items is
discarded if there are 15%or more defective items. What is the chance the shipment will be discarded?

Example 2 A manager of a company is interested int he percentage of customers that dislike his
company’s the new policy. How many customers should he sample to inquire about their opinion
in order to get an estimate for the percentage that is within .15 of the truth with probability 0.98?
If he samples 50 customers and the true percentage is 0.9 what is the chance that the estimate he
obtained from the sample is accurate within .15 of the true percentage.

Exercises: 7.71, 7.72, 7.73, 7.75, 7.77, 7.78, 7.80, 7.82, 7.85

You might also like