Scribd
Upload
28 views

Inference About One Population Variance: Outline

The document discusses inference about population variances. It outlines constructing a confidence interval and performing a hypothesis test for the population variance σ2. As an example, a 90% confidence interval is calculated for the variance of the lengths of steel shafts based on a sample. The technique of using the chi-squared distribution to test hypotheses about the population variance σ2 is also explained and an example problem is worked through.

Uploaded by

Chu Thuy Dung
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
28 views

Inference About One Population Variance: Outline

The document discusses inference about population variances. It outlines constructing a confidence interval and performing a hypothesis test for the population variance σ2. As an example, a 90% confidence interval is calculated for the variance of the lengths of steel shafts based on a sample. The technique of using the chi-squared distribution to test hypotheses about the population variance σ2 is also explained and an example problem is worked through.

Uploaded by

Chu Thuy Dung
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/ 10

03/09/2021

Lecture 2
Inference about One Population
Variance

Outline

• Chi-squared distribution

• Confidence interval for σ2

• Hypothesis testing for σ2

Inference About Population Variances

Inference About
Population Variances

Inference Inference
about σ2 about σ12/ σ22

1
03/09/2021

Inference about σ2

Quality control

• Quality: critical aspect of production


• A product is made of different components
• If components do not fit, product will not
function well
• Components should be made with as much
consistency as possible

Quality control
• Consistency means: there is as little variation
among the components as possible
• We measure variation by calculating the
variance
• Hence the need to make inference about
population variances
• One statistical technique is to construct a
confidence interval of the population
variance.

2
03/09/2021

Problem context

• Container-filling machines are used to package


a variety of liquids, including milk, soft drinks
and paint. Ideally, the amount of liquid should
vary only slightly, since large variations will
cause some containers to be underfilled
(cheating the customer) and some to be
overfilled (resulting in costly waste).

Problem context (cont.)

The director of a company that developed a


new type of machine boasts that this machine
can fill 1 litre (1000 cubic centimetre - cc)
containers so consistently that the variance of
the fills will be less than 1 cc. The statistical
technique that allows us to check this claim is
the hypothesis testing of the population
variance.

The technique

• To draw inference about variability, the


parameter of interest is the population
variance σ2.
• The sample variance s2 is an unbiased,
consistent and efficient point estimator for σ2.
• The inference about σ2 will be based on the
chi-squared sampling distribution

3
03/09/2021

The chi-squared sampling distribution

If the population is normally distributed and


has the variance , the statistic will
follow a distribution known as the chi-squared
distribution with (n-1) degrees of freedom.

The chi-squared sampling distribution


Characteristics of the chi-squared distribution:
Positively skewed
Range from 0 and ∞
Shape depends on number of degrees of
freedom
(http://www.lock5stat.com/StatKey/theoretical_distribution/theoretical_distribution.html#chi)

d.f. = 1

d.f. = 5 d.f. = 10

The chi-squared distribution

4
03/09/2021

Estimating the population variance

From the following probability statement


P(χ21-α/2 < χ2 < χ2α/2) = 1 – α
we have (by substituting χ2 = [(n – 1)s2]/σ2 )

The chi-squared table

α = .01
α =.01
1 - α =.99

χ21-α χ2.010
α
.990

χ2.01,10 = 23.2093

.995 .990 .975 .010 .005

Note that the chi-squared table in the textbook is not exactly


the same, but very similar to this one

Estimating the population variance

• The confidence interval for σ2 is

where 1 – α is the confidence level.

5
03/09/2021

Example

• A company manufactures steel shafts for use in


engines. One method for judging inconsistencies in
the production process is to determine the variance
of the lengths of the shafts.
• A random sample of 10 shafts produced the following
measurements of their lengths (in centimetres): 20.5,
19.8, 21.1, 20.2, 18.9, 19.6, 20.7, 20.1, 19.8, 19.0.
• Find a 90% confidence interval for the population
variance (assuming that the lengths of the steel
shafts are normally distributed)

Solution

• The problem objective is to construct a 90%


confidence interval for the parameter
• It is given that the lengths of the steel shafts are
normally distributed
• Therefore, the interval estimator for is:

Solution
• From the data we calculate:

• With 1 - α = 0.90, we have α/2 = 0.05

6
03/09/2021

Solution

• The 90% confidence interval estimate of


population variance of the lengths of the
shafts is (0.262, 1.34)

Another example

You randomly select and weigh 41 samples of


16-ounce bags of potato chips. The sample
standard deviation is 0.05 ounce.
Assuming the weights are normally distributed,
construct a 90% confidence interval for the
population standard deviation.

Answer

d.f. = n – 1 = 41 – 1 = 40

0.04 < σ < 0.06

The 90% confidence interval estimate of


population variance of the weight of bags of
potato chips is between 0.04 and 0.06 ounces.

7
03/09/2021

Testing the population variance

• Our null hypothesis is


H0: σ2 = σ02
• The alternative hypothesis is
HA: σ2 ≠ σ02 or HA: σ2 > σ02 or HA: σ2 < σ02
• The test statistic is

The test statistic has a χ2 distribution with d.f.


= n – 1, provided that the population is
normally distributed.

Solution to the problem in the


problem context
• The parameter to be tested is the population variance
• The population of 1-litre fills from the machine is
normally distributed
• Because we want to determine whether there is
enough evidence to support the claim, the alternative
hypothesis is:
Ha: <1
• The null hypothesis automatically follows as:
H0: =1

Solution to the problem in the


problem context
– The test:
H0: σ2 = 1
HA: σ2 < 1

Computations are shown next:

There is insufficient evidence


to reject the hypothesis that
the variance is equal to 1cc,
in favour of the alternative
hypothesis that it is smaller.

8
03/09/2021

Solution to the problem in the


problem context
α = 0.05 1 - α = 0.95

Rejection
region

13.8484 20.8

Do not reject the null hypothesis.

Exercise (Q11.11)
At the end of 2008, the variance in the semiannual
yields of overseas government bond was = 0.70.
A group of bond investors met at that time to discuss
future trends in overseas bond yields. Some expected
the variability in overseas bond yields to increase and
others took the opposite view. The following table
shows the semiannual yields for 12 overseas
countries as of March 6, 2009 (Barron’s, March 9,
2009).
Use = 0.05 to conduct the hypothesis test whether
the sample data indicate that the variance in bond
yields has changed from that at the end of 2008.

Exercise (Q11.11)

From the problem: 0.69, = 12, = 0.05

9
03/09/2021

Solution to Exercise (Q11.11)

1. The hypothesis of test:


: = 0.70
: ≠ 0.70
2. The test statistic: =
3. The rejection region is: > . !; = 21.920
or χ < .$%!, = 3.816
4. Value of test statistic: = 10.841
Do not reject

Another example
A company produces metal pipes of a
standard length, and claims that the
standard deviation of the length is at most
1.2 cm. One of its clients does not believe
and decides to test this claim by taking a
sample of 25 pipes and checking their
lengths. They found that the standard
deviation of the sample is 1.5 cm. Does this
provide evidence that the company’s claim is
incorrect?

Conclusion

• Chi – squared distribution.


• Confidence interval for population variance.
• Chi – squared test for population variance.

10

You might also like