Confidence Intervals

Estimating a Population Mean: σ Known

Requirements for Estimating μ when σ is known

  1. The sample must be a simple random sample.
  2. The value of σ is known
  3. Either the population is already normal or n ≥ 30 so the Central Limit Theorem can be applied.

Terms:

Point Estimate –

Confidence Interval –

Confidence Level –

α

1 – α

Critical Values

zα/2 is the symbol used to denote

Therefore,

Margin of Error

Sample Size for Estimating μ

Examples

1.Find the critical values for 90%, 95%, 98%, and 99%.

2.The National Assessment of Educational Progress includes a short test of quantitative skills, covering mainly basic arithmetic and the ability to apply it to realistic problems. This test was given to 1000 women of ages 21 to 25 years. Their mean quantitative score was 275. Suppose that you know that σ is 58. Give a 95% confidence interval for the mean score in the population of all young women ages 21 to 25 years. Interpret the resulting interval in words that a statistically naïve reader would understand.

3.The U.S. Department of Agriculture wants to determine the average number of eggs that children under 14 years of age consume each year. A random sample of 900 such children is obtained. In this sample, the average number of eggs consumed was = 92. Assume that

σ = 16.

a)Find a 95% confidence interval for the mean.

b)Suppose the department was hoping that µ = 95. What would you tell them based on your interval?

4.A professor wants to know the average age of the night students at a certain college. The variance (σ2) is known to be 36. A random sample of 400 night students yields a sample mean of 27 years.

a)Find a 90% confidence interval for the mean.

b)What is the width for a 99% confidence interval for the mean?

5.Starting annual salaries for college graduates with business administration degrees are believed to have a standard deviation of approximately $1,700. How large a sample should be taken if we want to be 95% confident that the maximum sampling error is $500?

Estimating a Population Proportion – Section 3.3

Requirements for Using a Normal Distribution as an Approximation to a Binomial Distribution

  1. The sample must be a simple random sample.
  2. The conditions for the Binomial Distribution are satisfied.
  3. Both np and nq are greater than or equal to 5.

Notation

p

(can be in proportion, percent or probability form)

Terms:

Point Estimate

Margin of Error

Sample Size for Estimating Proportion p

Examples

1.Use the confidence interval (0.444, 0.484) to find the point estimate and the margin of error.

2.A private consulting firm is given a governmental grant to study underreporting of income on federal income tax returns. The firm takes a random sample of 160 taxpayers and determines that 64 of them underreported their income on their tax return.

a)What is the population proportion pin this setting?

b)What is the sample proportion of taxpayers underreporting their income on their tax return?

c)What is the standard deviation for this proportion?

d)Construct a 95% confidence interval for this population proportion.

e)Interpret the resulting interval from part (d) in words that a statistically naïve reader would understand.

3.In a random sample of 100 students at a particular college, 60 indicated that they favored having the option of receiving pass-fail grades for elective courses. Obtain a 95% confidence interval for the proportion of the population of students who favor pass-fail grades for elective courses. Does this confidence interval contain the value p = .5? Explain why this particular value might be of interest.

4.Suppose you want a margin of error of 0.060 with aconfidence interval of 99% and that you know from a previous study that. What sample size should you use?

Estimating a Population Mean: σNot Known – Section 3.4

Requirements for Estimating μ when σ is not known

  1. The sample must be a simple random sample.
  2. Either the population is already normal or n ≥ 30 so the Central Limit Theorem can be applied.

Terms:

Point Estimate –

Student t distribution

The use of the t distribution is based on the following argument. If the population is normally distributed with mean  and variance 2, the sample mean is normally distributed (assuming the CLT applies) with mean and standard deviation and the standardized Z score,, is a standard normal variable. If  is unknown and replaced by the sample standard deviation s in the Z score formula, the Z score no longer follows the standard normal distribution. If we replace the true standard deviation by its estimate s, we obtain what is called the t score, wheret =

This t score follows what is called the Student t distribution.

The statistic t can be thought of as an estimated standardized normal variable because the estimated standard deviation s is used instead of  . The larger the sample becomes, the more the t becomes a normal.

Characteristics of the t Distribution

1.The Student t distribution is different for different sample sizes.

2.It is symmetric about 0 and ranges from –  to .

3.The distribution is bell shaped and has approx. the same appearance as the standard

normal.

4.The mean is 0.

5.The t distribution depends on a parameter  (nu) called the degrees of freedom of the

distribution. When constructing confidence intervals for the population mean, the

appropriate number of degrees of freedom is  = (n –1), where n is the sample size.

6.The variance of the t distribution is for  >2

7.The variance is always > 1.

8.As  increases, the variance of the t distribution approaches 1 and the shape approaches

that of the standard normal distribution.

9.Because the variance of t exceeds 1, the t distribution is slightly flatter in the middle

than the standard normal distribution and has thicker tails.

To use the table, we must know the degrees of freedom and how much area we want in the right tail of the curve.

The symbol t/2, denotes the value of t such that the area to its right is /2 and t has  degrees of freedom. The value t/2, satisfies the equation P(tt/2,) = /2 where the random variable t has the t distribution with  degrees of freedom (d.f.).

Conditions for Using the Student t distribution

  1. σ is unknown
  2. Either the population has a distribution that is essentially normal or n ≥ 30.

Confidence Interval –

Margin of Error

Examples

1.A company has a new process for manufacturing large artificial sapphires. The production of each gem is expensive, so the number available for examination is limited. In a trial run, 12 sapphires are produced. The mean weight for these 12 gems is= 6.75 carats, and the sample standard deviation is s = 0.33 carats. Assuming the population distribution is normal, find a 95% confidence interval for the mean weight for this type of artificial sapphire.

2.The life in hours of a 75-watt light bulb is known to be normally distributed with a standard deviation of σ = 25 hours. A random sample of 20 bulbs has a mean life of =1014 hours. Construct a 95% confidence interval for the mean life.

Estimating a Difference in Population Means: σ Known

Requirements for Estimating μ1 – µ2 when σ is known

  1. The two samples must be INDEPENDENT simple random samples.
  2. The value of σ is known for each of the populations.
  3. Either each population is already normal or n ≥ 30 for both samples so the Central Limit Theorem can be applied.

Terms:

Point Estimate –

Confidence Interval –

Margin of Error –

Example

It is desired to compare the average ages at which men and women get their first driver’s license. A random sample of 75 men yielded a mean of 17.3 years and a standard deviation of 4.7 years. A random sample of 96 women yielded a mean of 19.6 years and a standard deviation of 5.1 years.

a)Construct a 95% confidence interval for the difference between population means (i.e. for (μ1 - μ2)). Let the men represent population 1.

b)Based on your confidence interval in part (a), can you conclude that there is a difference between the mean age at which males and females get their first license? Why or why not? If there is a difference, which sex gets their license earlier? (Be as specific as possible.)

Estimating a Difference in Population Proportions – Section 3.5

Requirements for Estimating p1 – p2

  1. The two samples must be INDEPENDENT simple random samples.
  2. The conditions for the Binomial Distribution are satisfied for both populations.
  3. Both np and nq are greater than or equal to 5 for both groups.

Terms:

Point Estimate –

Confidence Interval –

Margin of Error –

Example

A city planner claims that homeowners tend to have closer ties to their community than do renters. Thus, homeowners are more willing to pay for good schools and recreational facilities than are renters. In a random sample of 120 homeowners, 50 stated that the local tax rates were too high and the remaining 70 state that tax rates were “about right”. In an independent random sample of 200 renters, 70 thought that the local tax rates were too high and 130 thought they were “about right”.

a)Find a 99% confidence interval for the difference in the population proportions who think taxes are too high.

b)Does the data support the city planner’s claim? Support your answer.