Table of Contents Table of Contents
Previous Page  492 / 762 Next Page
Basic version Information
Show Menu
Previous Page 492 / 762 Next Page
Page Background

492

Chapter 11: Confidence Intervals – Large Samples

As the sample sizes

and

increase, the shape of the distribution of

the differences of the sample proportions obtained from any population

(distribution) will approach a normal distribution.

The distribution of the differences of the sample proportions will have a

mean given by

̂

̂

.

The distribution of the differences of the sample proportions will have a

standard deviation given by

̂

̂

, where

and

are the respective population proportions of interest.

These properties can aid us in the construction of a (1 -

)

100 percent

confidence interval for the difference of two population proportions. Again,

since we do not know the values of the true proportions, we will use the

corresponding point estimates for these true proportions.

The general equation used in constructing a (1 -

)

100 percent confidence

interval for the difference between two population proportions for large

samples, that is when

5,

5, and

5, is given next.

Example 11-10:

A study was conducted to determine whether remediation

in basic mathematics enabled students, who did not have the necessary ACT

requirements, to be more successful in a college algebra course. Success here

means a student received a grade of C or better in the college algebra course

and remediation was for one-year prior to the college algebra course.

Table

11-1

shows the results of the study.