Estimate Difference of Proportions from Summary Statistics

How To... > Work with Statistical Objects > Estimate Parameters and Compute Confidence Intervals > Estimate Difference of Proportions from Summary Statistics

_img489

Sometimes you don’t have the raw data from which you wish to estimate a difference of proportions. Let’s take a concrete example in which you know the counts for sex and job classification (salaried or hourly) of 42 employees and you wish to estimate any difference in proportion of males between the two job classifications. The summary data are as follows:

Salaried, female: 10

Salaried, male: 12

Hourly, female: 17

Hourly, male: 13

1.	Create a new estimate by dragging one from the shelf or by choosing Object \| New \| Interval Estimate.

_img154

2.	From the pop-up menu in the estimate’s upper-right corner, choose Difference of Proportions.

3.	Double-click the bold, blue field FirstAttribute in the body of the estimate, and type a reasonable name, such as Salaried.

4.	Press Tab to move to the next field, and continue inputting appropriate values.

As you enter values, the text changes throughout the object. The result for this example is shown here.

You can change the confidence level by double-clicking it and typing a different value.

Note that the estimate of the difference of proportions goes from –0.16 to 0.38 at the 95% confidence level. Because this confidence interval includes zero, we are not justified in saying that the difference of proportions is greater than zero.