Analysis of Data Using SDA [Edited 11/02/2007 4:10pm ADA]:

(1) Open the General Social Survey Datafile "http://sda.berkeley.edu/cgi-bin32/hsda?harcsda+gss06".  A page with the following will be displayed: 

NOTE: the following example is an introduction for the most frequent occuring social science analysis with two variables that have categories for answers. For the following example "relig" has specific categories (PROTESTANT, CATHOLIC, JEWISH, NONE, OTHER) and "abany" also has specific categories (YES, NO).  If one of your variables was continous such as AGE (years), you should use an alternative type of analysis possibly "Comparison of means".  This is beyond the purpose of this tutorial

Start SDA for Crosstab Page

(2) Type in your analysis choices: Click beside "Frequencies or Crosstabs" then click the Start button.  The "SDA Tables Program" page will open:

  1. Type the name of your dependent variable, abany from the GSS in this case, in the Row  box
  2. Type the name of  your independent variable, relig, in the Column  box
  3. Type "year (1996)"  in the Selection Filter(s):  box.  This will choose the GSS for the 1996 data, leave blank if you want the complete 1972-2006 data. You also could choose 2006 or any other another year. I choose 1996 because the religious categories were simpler and more appropriate to demonstrate on a class projector.
  4. Click the down arrow by Weight and choose No Weight
  5. Click the Percentage  "column" box
  6. Click by Statistics and choose 2 decimals
  7. Click the down arrow by Type of chart and choose [no Chart]
  8. Click the Run the Table button
(3) Your table(s) and statistics will be displayed.  There will be two tables. The first table displays information for all people who took the survey.  The second table, Statistics for all valid cases, displays the data for just those who answered the questions you are examining.  In general you would use the Statistics for all valid case unless you have an interest in those who did not answer.
 
 Note: First Print your table and statistics output.  If part of the output page is cut off: 
(1) Decrease the print size
Microsoft Explorer: Choose Text Size from the View menu and select a smaller size
Netscape: Choose from the View menu Decrease font. You may have to repeat this several times to get a smell enough taple to print. 
and/or
(2) Change the print option to Landscape


(3) Reverse the column, row and percentage choices and run the table again

(4) Consider comparing diffirences in categories of your independant variable if the data is appropriate  (means are appropriate for data like money, weight, temperature but not classifications like ethnic groups, religious groubs, etc.

If your table still has two many categories you can change your hypothesis and use different  variables or you may want to consider recoding  your variables or obtaining a different type of output such as a t test.  This is more advanced but for those who want to try Advanced analysis techniques using SDA
(a) At the top of your page will be general descriptions of the variables
Variable Information Header
The range of the variables, and filters in selection for year are listed
(b) The distribution table for your variables all people taking the survey appears next

Distribution Table

This table lists the distribution of your dependent variable across your independent variable.  Check reading a table for specifics on how to interpret this table
(c)  The last information on this page is the statistics and summary information about your table

Table Statistics

The Summary Statistics information contains a wide range of Statistics about your table.  Choose statistics that you know and apply to this data. Check Interpreting Statistics to find out more about these statistics
Advanced analysis techniques using SDA

Back


  HELP