 |
How do I stratify results by one or more variables?
SEER*Stat results will be displayed or stratified by the format groups defined in the variables
listed in the row, column, and page dimensions of the Table Tab. Therefore, use two mechanisms in SEER*Stat to define how your results are stratified:
- Using the dictionary, define format groupings
for the variables of interest. For example,
you may want to show frequencies for cases diagnosed
in two 5-year periods (2000-2004 vs 1995-1999). In the
dictionary, you would need to edit the "Year of Diagnosis" variable
and create two groupings with the appropriate labels
and values.
- Include the stratification variables in a dimension of the Table Tab. In our example, you would
want to put your user-defined year of diagnosis variable in the row, column, or page dimension.
Example: Stratify Frequencies by Sex
Calculate the number of malignant primary tumors in the
SEER 17 registries diagnosed in 1973-2004. Start by opening
a new Frequency session. Select the SEER 17 registry database
on the Data Tab. Malignant tumors and known ages will be selected by default
(due to the "Select only " check boxes at the top of the Selection Tab).
To stratify results by sex, go to the Table Tab. Use the
'+' next to the "Race, Sex, Year Dx, Registry, County" category
in the list of
"Available Variables" to expand the list of variables. Select
the "Sex" variable, then use one of the available
buttons to the right (Row, Column, Page) to add sex to the
list of display variables. When you generate the results,
the Sex variable is displayed by the three groupings defined
for this variable in the SEER database: 'Male', 'Female',
and 'Male and Female'.
Change Variable Groupings
A unique dictionary is associated with each SEER*Stat database containing one pre-defined, formatted
variable for each data field. You can create your own formats for a field's values by creating new
variables ("user-defined" variables). The variables distributed with a database are called "standard
variables" and have pre-formatted groupings for common use. To view the definition of any variable:
- Open the dictionary by selecting Dictionary from the File menu,
using the
 on the toolbar; or
double-clicking on the variable in the Available Variables
box or the Display Variables box on the Table Tab.
- Double-click the variable in the Dictionary
window, or use the Create button to open the Edit
Variable window and see the variable's
definition.
- Groupings are essentially format statements that allow you to label individual or
groups of values. When you click on a label in the Groupings box, the values
associated with the label will be highlighted in the Values box. All values occurring
in the database for the variable are listed in the Values box. The list of values can not be changed.
To change the groupings for any variable in the database, you need to create a new variable, called
a "user-defined" variable. Let's say you did not want to include the grouping for both sexes combined in
the output of the example above. You would need to define a user-defined variable with just two
groupings, "Male" and "Female", using the values available in the standard sex variable. This could be done
by deleting the "Male and Female" grouping and saving the variable with a new name. Most of the tutorials
include steps which provide detailed instructions for creating user-defined variables.
Setting Variables on the Table Tab
The dimensions (row, column, page) available on the Table Tab vary in the different sessions. In most sessions, you can
choose the dimension to which the variable is assigned. A separate row, column, or page will be created
for each of the variable's format groupings. The dimensions available in each type of
session are:
- Frequency: Page, Row, Column
- Rate: Page, Row, Column
- Survival: Page, Row, Column (for Summary pages)
- Limited-Duration Prevalence: Page, Row, Column
- MP-SIR: Page, Row, Column
- Case Listing: Column
|
