Analysis of Contingency Tables
This applet will allow you to vary the cell entries in a
contingency table and computes chi-square and its associated
probability value. It also shows the contribution of each cell by
varying the brightness of the cell colors.
The applet is shown below, where the data are taken from Table
9.2, which relates to the Darley and Latane study of helping behavior
as a function of the number of bystanders.
Enter the appropriate values
in the cells. Blue cells represent higher than expected numbers
of observations, while pink cells represent fewer observations
than expected. The brighter the colors, the greater the
difference from expected. |
|
- Enter the data from Pugh's experiment on convictions for
rape.
- Enter the data from Latane and Dabbs in Section 19.6.
- For the Latane and Dabbs study, where could you add 10
observations to make the most difference in chi-square?
- If you double the total sample size without making changes to
any of the cells, the applet will automatically double those entries.
Do this and observe the changes in chi-square and the color of the
cells.
- What does the last item say about power and sample size?
- What if you just sampled twice as many observations in one of
the rows?
Comments to: Gary.McClelland@Colorado.e
du
Return to index