As I help course members with their final projects I'll add notes about the various Statistica maneuvers necessary.
Recoding data. Suppose you have survey data from four settings (e.g., colleges of different type), and -- discovering that you don't have enough subjects in one or more of the settings to perform valid statistics -- you decide to recode the data into fewer categories. Do the following.
1. Double-click the variable name in Statistica's spreadsheet window. A window
opens with various aspects of the variable listed
Click the Text Values window to see a listing of the labels for the values of "COLLEGE." Note the four numeric codes for "bico," "libart," "privat," and "public" respectively. Suppose you decide to group the first two and the last two types of institution in a new two-way "small"-"big" categorization by recoding each "2" as "1" and each "4" as "3." Start by clicking on "bico" and changing the text field to "small." Then click on "privat" and change it to "big." Click OK.
2. Now, from the main spreadsheet window, click the Vars button near the top of the screen. Select the last option, Recode, from the drop-down menu. In the First Include If: field under Category 1, enter the instruction "v4=2," that tells Statistica to recode all cases of variable 4, "COLLEGE," with value 2, "libart," as New Value 1 (the default). Then enter "v4=4" in the Category 2 field and "3" in the New Value 2 field to recode the "public" cases. Click OK (Recode), and confirm with Yes in the next popup window and No in the last, Save Conditions?, window. You should now see only "small" and "large" as values of COLLEGE in the spreadsheet.
3. At this point it would be a good idea to save your file to a local folder under an informative name, e.g., "p205.02.survey1.recoded."
Now you're ready to use the new COLLEGE variable as a sorting (independent) variable in Crosstabulation, T-test, and/or ANOVA.