Data:  K:\nd.edu\user22\yli\Public\593sp01\Data\loglin.dta

Using loglinear regression method, analyze the effects of race and sex on high school dropout.  Try and compare at least four different models (e.g., saturated, joint independence, etc.).  Interpret the results in your best model.

Note: You may need to switch to Stata 5.0 for this exercise if you want the "loglin" command to give you output on the Goodness of Fit Chi-square.  Stata 6.0 does not show the Goodness of Fit statistics in loglinear model results.