Thus far, we have used the DEFINE statement to change the appearance of our reports, doing such things as specifying the format of a variable, changing the width of a column, specifying column labels, and changing the justification of a column. The truth is that the DEFINE statement is even more powerful when it is used to tell SAS how you want SAS to use each variable that appears in your report. There are six possible usages — DISPLAY, ANALYSIS, ORDER, GROUP, ACROSS, and COMPUTED — which we'll investigate in the course of the next five sections of this lesson. By default, the REPORT procedure uses:

• character variables as display variables
• numeric variables as analysis variables, which are used to calculate the SUM statistic

Because we haven't explicitly defined any variable usages, the reports we have created thus far have contained only display and analysis variables. For example, here's the most recent report we created:

Because the second, fourth, and fifth columns, Year, Par, and Yards, are numeric variables, SAS treats them as analysis variables by default. Analysis variables are used to calculate a statistic ... by default a SUM.

Now, because the first and third columns, Name and Type, are character variables, SAS treats them as display variables by default. What impact does that have on the appearance of the report? It means three things: 1) A report that contains one or more display variables necessarily contains a detail row for each observation in the data set, 2) Each detail row contains a value for each display variable, and 3) The order of the rows in the report matches the order of the observations in the data set.

We can alter the order in which the rows appear in our report by defining an order variable (or two or three ...). An order variable orders the detail of the rows in a report according to the variable's formatted values. Let's take a look at an example.