Database Reference
In-Depth Information
Step 12: Create the Two-Way Chi-Square Test
For the next type of statistical analysis, we want to show whether different order methods gener-
ate more promotional participation by retailers.
1.
In the
Explorer Bar
, mouse over the
Page Explorer
tab and select the
Report Pages
folder.
2.
In the work area, copy and paste the
Box Plot Chart
page.
3.
With the
Box Plot Chart1
page selected, double-click the
Render Variable
property in
the
Properties
pane.
The Render Variable dialog box is displayed.
4.
In the
Render for
pane, remove the check for option
3
.
5.
Add a check for option
4
.
6.
Click
OK
to close the dialog box.
7.
In the
Properties
pane, change the
Name
property to
Two-Way Chi-Square Analysis
.
8.
In the
Report Pages
pane of the work area, double-click the
Two-Way Chi-Square
Analysis
page to open the page design.
9.
In the page body, click the
Boxplot
object to select it.
10.
From the toolbar, click the
Cut
button to remove the data container.
11.
From the
To o l b o x
tab of the
Insertable Objects
pane, drag a
Statistics
object into the
page body.
The Select Statistic dialog box is displayed.
12.
In the left pane of the dialog box, expand
Nonparametric Tests
.
13.
Click
Two-Way Chi-Square Test
to select it.
14.
Click
OK
to close the dialog box.
A Create Two-Way Chi-Square Test dialog box is displayed.
15.
Select the
Two-Way Chi-Square with counts measure
radio button.
16.
Click
OK
to close the dialog box.
17.
In the work area, click the
Two-Way Chi-Square Test
container to select it.
18.
In the
Properties
pane, change the
Query
property to
Statistics_query
.
19.
Change the
Show observed count
property to
No
.
20.
Change the
Show percent of row
property to
Ye s
.
This will allow the crosstab to show the percentage that each order method contributes
to the number of orders for a campaign.
21.
From the
Data Items
tab of the
Insertable Objects
pane, drag the following items into
the statistics container: