3.0 Web Investigator

3.7 Run a Frequency or Table button

From this button the user can run simple statistics such as frequencies (i.e., basic survey results) and cross-tabulations on tag set data. The setup menu shown below appears when the button is pressed; all but the bottom field, however, have the word 'Chose' in them.

Using the variables from our tag set for the frequency table, we are going to set up a simple, two-way table of variable x by variable y, that is R0214700 (race) by R0214800 (sex). This will produce a single table with frequencies in each of 6 cells, 3 for each category of race multiplied by the 2 categories for sex.

Here are the steps to go through to set up a simple frequency.

Tabulation Options Step 1: Select a Row Variable (required)

Select a specific question in the survey that you are interested in analyzing. This is called the "row" variable because the response categories for this question will appear as rows of a table.

To continue our example,

Select R0214700 (R'S RACIAL/ETHNIC COHORT...) from the tag set by using the drop-down selection box.

You may browse through the drop-down list of all the tagged variables (by 'name', 'variable title' and 'year') in the tag set.

Tabulation Options Step 2: Select a Column Variable (optional)

A "column" variable is an optional second question to be analyzed together with the row variable. Its response categories will appear as columns of a table.

To continue our example,

Select R0214800 (SEX OF R) from the tag set by using the drop-down selection box.

A column variable is often used to analyze the results of a question (the row variable) by demographic groups, but it may also be used to "cross-tabulate" any two variables.

Note: In a cross-tab analysis, the column variable is sometimes also referred to as the "independent" variable. When running cross-tabs, it is conventional to select the independent variable as the "column variable," and the "dependent variable" as the "row variable."

That's it for the 2-way frequency table.

Click on Submit Table.

When you click on the submit button, you will be directed to you account and you will see that the file is being created (running). After the table has been created in your account, you can view it in text of html format in you default browser from there. The 'text' file will look like the following window.

Universe Restrictors Step 3: Select Condition 1 (optional)

We could of course have excluded one of the sexes in the table we just created. Choosing a conditional variable will cause the analysis to further limit the universe of variables to only those that meet the condition expressed. A conditional variable can be used whether or not a column variable was chosen.

To continue with our example, let's limit the sex to females only

Make the same selections (row and column) as in Steps 1 & 2 above; then Select 'SEX OF R' and set it '=' to '2' without the quotes; and Click on Sumbit Table.

The following results will be produced.

Universe Restrictors: Select Condition 2 (optional)

A second conditional variable is available to use whether or not a column variable was chosen. This second conditional variable will further limit the universe of available variables in accordance with the Conditional Operator (required) chosen.

Tabulation Options: Select Page Variable (optional)

A page variable, a.k.a. control variable, can be used whether or not a column variable was chosen.

When a control variable is selected along with a column variable, the output will be separate cross-tabulation tables for each response category of the control variable or a three-way table for variable x by variable y by variable z.

In our example, variable x = RACE and variable y = SEX, in a cross-tabulation, you could add R0010600 S02Q01 MARITAL STATUS 1979 to the tag set, then select it as the "control variable", i.e., variable z, in order to view the complete sample results of RACE by SEX as well as the results broken down into separate tables for each of the five marital status response categories: PRESENTLY MARRIED; WIDOWED; DIVORCED; SEPARATED; NEVER MARRIED-ANNUL.

Of course, one could limit the control variable in the above example to only those cases in which the marital status was equal to 'divorced' (or response value = '3' ). The following would be the result: Divorced by Race and by Gender.

Try to recreate the table above. Hints:

Keep the tag set of four variables (4_common_vars) open. Add the filter "R0010600 MARITAL STATUS 1979" using the Word in Title index. Review Tagged Variables and Save Tag Set with R0010600 in it as "5_common_vars". Open the new tag set, 5_common_vars.html and Run a Frequency or Table. View the frequency cross-tabulation from your account.

Tabulation Options: Select Weight Option (optional)

The weight option will inflate the cell frequencies by the weight selected. These are not the custom weights found under the Custom Weights button, but are the single-year weights produced for each round. The weights do not have to be included in the tag-set because the web investigator includes all the weights for all rounds in the drop down menu.

Go to 3.0 Web Investigator.

Go to 3.6 Extract Tagged Variables.

Go to 3.8 Code Book.

Go back to the Table of Contents.