jingju11 data dset; input gender $ pass n; datalines; m 1 137 m 0 283 f 1 24 f 0 71 ; proc freq data=dset; tables gender*pass/chisq; weight n; run;
xjuchenwei 这是SAS分析的结果 The SAS System 22:52 Saturday, June 24, 2009 1 The FREQ Procedure Table of gender by pass gender pass Frequency? Percent ? Row Pct ? Col Pct ? 0? 1? Total 儍儍儍儍儓儍儍儍儍垉儍儍儍儓 f ? 71 ? 24 ? 95 ? 13.79 ? 4.66 ? 18.45 ? 74.74 ? 25.26 ? ? 20.06 ? 14.91 ? 儍儍儍儍儓儍儍儍儍垉儍儍儍儓 m ? 283 ? 137 ? 420 ? 54.95 ? 26.60 ? 81.55 ? 67.38 ? 32.62 ? ? 79.94 ? 85.09 ? 儍儍儍儍儓儍儍儍儍垉儍儍儍儓 Total 354 161 515 68.74 31.26 100.00 Statistics for Table of gender by pass Statistic DF Value Prob 儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍 Chi-Square 1 1.9508 0.1625 Likelihood Ratio Chi-Square 1 2.0118 0.1561 Continuity Adj. Chi-Square 1 1.6235 0.2026 Mantel-Haenszel Chi-Square 1 1.9471 0.1629 Phi Coefficient 0.0615 Contingency Coefficient 0.0614 Cramer's V 0.0615 Fisher's Exact Test 儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍儍 Cell (1,1) Frequency (F) 71 Left-sided Pr <= F 0.9375 Right-sided Pr >= F 0.1001 Table Probability (P) 0.0376 Two-sided Pr <= P 0.1789 Sample Size = 515
hexm26 就是用2x2关联表,7楼给出背景知识,6楼给出程序,10楼给出输出结果,应该是最合理的解答。 4楼,5楼,8楼,12楼,15楼,16楼给于忽视。 17楼的提出很有意思的东西,关联表分析本质上就是卡方分析,而二进制逻辑回归分析用最大似然法估计出来的应该和卡方分析的结论一致。