/********************************************************************* * STAT 7030 - Categorical Data Analysis * Peng Zeng (Auburn University) * 2024-2-6 *********************************************************************/ data admission; input dept $ gender $ yes no @@; total = yes + no; datalines; anth female 32 81 anth male 21 41 astr female 6 0 astr male 3 8 chem female 12 43 chem male 34 110 clas female 3 1 clas male 4 0 comm female 52 149 comm male 5 10 comp female 8 7 comp male 6 12 engl female 35 100 engl male 30 112 geog female 9 1 geog male 11 11 geol female 6 3 geol male 15 6 germ female 17 0 germ male 4 1 hist female 9 9 hist male 21 19 lati female 26 7 lati male 25 16 ling female 21 10 ling male 7 8 math female 25 18 math male 31 37 phil female 3 0 phil male 9 6 phys female 10 11 phys male 25 53 poli female 25 34 poli male 39 49 psyc female 2 123 psyc male 4 41 reli female 3 3 reli male 0 2 roma female 29 13 roma male 6 3 soci female 16 33 soci male 7 17 stat female 23 9 stat male 36 14 zool female 4 62 zool male 10 54 ; proc print data = admission; run; proc genmod data = admission; class dept; model yes / total = dept / dist = bin link = logit r; output out = residuals stdreschi = streschi; run; proc sort data = residuals; by gender; proc print data = residuals; run; /*** logistic regression with three departments deleted ***/ data refined; set admission; if dept in ('astr', 'geog', 'psyc') then delete; proc print data = refined; run; proc genmod data = refined; class dept; model yes / total = dept / dist = bin link = logit; run; /*** logistic regression using dept and gender as predictors ***/ proc genmod data = admission; class dept gender; model yes / total = dept gender / dist = bin link = logit; run; /********************************************************************* * THE END *********************************************************************/