|
Help (wait for the button to appear above) |
|
||||
The Paired item on the T Statistics submenu on the Stat menu provides confidence intervals and/or hypothesis tests for a difference in population means with paired data. The results are displayed in the Results panels.
In the dialog box, select the two variables to be analyzed. This procedure is illustrated using the Labor Force data, which gives the rate of participation of women in the labor force for 19 American cities in two years, 1968 and 1972. The differences are computed by subtracting the variable chosen second from that chosen first. By checking the Save differences checkbox, these differences will be stored in a new column in the data table.
Click the Next -> button to select either a confidence interval (specifying confidence level) or a hypothesis test (specifying null mean or mean difference and alternative hypothesis). In the Labor Force example, it is of interest to test whether the rate of women participating in the labor force increased between 1968 and 1972. Since the 1968 observations are subtracted from their 1972 counterparts, an upper tailed alternative is selected.
Click on the Calculate button and the output is displayed in the Results panel. In the hypothesis testing situation, the output consists of the null hypothesized difference in means ("Delta0"), the estimate of the difference, its standard error, number of degrees of freedom, the T test statistic value, and the associated p value. In the confidence interval case, the output includes the estimated difference, its standard error, the number of degrees of freedom, and upper and lower confidence limits for the difference.
In this case there is strong evidence (p = .0122) that labor force did actually increase.