Split the variable EDUCAT in the depression data set into two groups, those who completed high school and those who did not. Also split INCOME at 18 or less and greater than 18 (this is thousands of dollars). Run a log-linear analysis of SEX versus grouped EDUCAT versus grouped INCOME and describe what associations you find. Does the causal ordering of these variables suggested in paragraph 6 of Section 17.13 help in understanding the results?