MTB > plot c2 c1 1.05+ - * * * * 2 * * 2 * C2 - - - 0.70+ - - - - 0.35+ - - - - 0.00+ 2 * 2 * * * * * * * * * - --+---------+---------+---------+---------+---------+----C1 5.0 10.0 15.0 20.0 25.0 30.0 MTB > regress c2 1 c1 c11 c12 The regression equation is C2 = - 0.092 + 0.0315 C1 Predictor Coef Stdev t-ratio p Constant -0.0922 0.1833 -0.50 0.620 C1 0.031528 0.009606 3.28 0.003 s = 0.4271 R-sq = 31.9% R-sq(adj) = 28.9% Analysis of Variance SOURCE DF SS MS F p Regression 1 1.9648 1.9648 10.77 0.003 Error 23 4.1952 0.1824 Total 24 6.1600 Unusual Observations Obs. C1 C2 Fit Stdev.Fit Residual St.Resid 2 29.0 0.0000 0.8221 0.1444 -0.8221 -2.05R 25 8.0 1.0000 0.1600 0.1207 0.8400 2.05R R denotes an obs. with a large st. resid. MTB > print c1 c2 c12 ROW C1 C2 C12 1 14 0 0.349199 2 29 0 0.822122 3 6 0 0.096973 4 25 1 0.696009 5 18 1 0.475312 6 4 0 0.033916 7 18 0 0.475312 8 12 0 0.286142 9 22 1 0.601425 10 6 0 0.096973 11 30 1 0.853651 12 11 0 0.254614 13 30 1 0.853651 14 5 0 0.065444 15 20 1 0.538368 16 13 0 0.317670 17 9 0 0.191557 18 32 1 0.916707 19 24 0 0.664481 20 13 1 0.317670 21 19 0 0.506840 22 4 0 0.033916 23 28 1 0.790594 24 22 1 0.601425 25 8 1 0.160029 MTB > mplot c2 c1, c12 c1. 1.05+ - A A A A 2 A A 2 A - B - B 2 - B 0.70+ B - 2 B - B - 2 B - 0.35+ 2 B - B B - B - B - B 2 0.00+ 4 A 2 A A A A A A A A A - --+---------+---------+---------+---------+---------+---- 5.0 10.0 15.0 20.0 25.0 30.0 A = C2 vs. C1 B = C12 vs. C1 ****start weighted least squares procedures******************************** MTB > name c3 'wi' MTB > let c3 = 1/(c12*(1 - c12)) MTB > print c1 c2 c12 c3 ROW C1 C2 C12 wi 1 14 0 0.349199 4.4003 2 29 0 0.822122 6.8382 3 6 0 0.096973 11.4196 4 25 1 0.696009 4.7263 5 18 1 0.475312 4.0098 6 4 0 0.033916 30.5196 7 18 0 0.475312 4.0098 8 12 0 0.286142 4.8956 9 22 1 0.601425 4.1717 10 6 0 0.096973 11.4196 11 30 1 0.853651 8.0044 12 11 0 0.254614 5.2691 13 30 1 0.853651 8.0044 14 5 0 0.065444 16.3502 15 20 1 0.538368 4.0237 16 13 0 0.317670 4.6135 17 9 0 0.191557 6.4573 18 32 1 0.916707 13.0967 19 24 0 0.664481 4.4854 20 13 1 0.317670 4.6135 21 19 0 0.506840 4.0007 22 4 0 0.033916 30.5196 23 28 1 0.790594 6.0403 24 22 1 0.601425 4.1717 25 8 1 0.160029 7.4394 MTB > regress c2 1 c1 c21 c22; SUBC> weights c3. The regression equation is C2 = - 0.117 + 0.0327 C1 Predictor Coef Stdev t-ratio p Constant -0.1171 0.1118 -1.05 0.306 C1 0.032672 0.006644 4.92 0.000 Analysis of Variance SOURCE DF SS MS F p Regression 1 23.809 23.809 24.18 0.000 Error 23 22.642 0.984 Total 24 46.451 Unusual Observations Obs. C1 C2 Fit Stdev.Fit Residual St.Resid 2 29.0 0.000 0.830 0.124 -0.830 -2.32R 6 4.0 0.000 0.014 0.092 -0.014 -0.09 X 18 32.0 1.000 0.928 0.141 0.072 0.30 X 22 4.0 0.000 0.014 0.092 -0.014 -0.09 X 25 8.0 1.000 0.144 0.077 0.856 2.41R R denotes an obs. with a large st. resid. X denotes an obs. whose X value gives it large influence. MTB > print c1 c2 c12 c22 ROW C1 C2 C12 C22 1 14 0 0.349199 0.340294 2 29 0 0.822122 0.830373 3 6 0 0.096973 0.078918 4 25 1 0.696009 0.699685 5 18 1 0.475312 0.470981 6 4 0 0.033916 0.013574 7 18 0 0.475312 0.470981 8 12 0 0.286142 0.274950 9 22 1 0.601425 0.601669 10 6 0 0.096973 0.078918 11 30 1 0.853651 0.863045 12 11 0 0.254614 0.242278 13 30 1 0.853651 0.863045 14 5 0 0.065444 0.046246 15 20 1 0.538368 0.536325 16 13 0 0.317670 0.307622 17 9 0 0.191557 0.176934 18 32 1 0.916707 0.928388 19 24 0 0.664481 0.667013 20 13 1 0.317670 0.307622 21 19 0 0.506840 0.503653 22 4 0 0.033916 0.013574 23 28 1 0.790594 0.797701 24 22 1 0.601425 0.601669 25 8 1 0.160029 0.144262 *********fit from logistic BMDPLR********** MTB > let c7 = expo(-3.06 + .161*c1) MTB > let c8 = c7/(1 + c7) MTB > print c1 c2 c8 ROW C1 C2 C8 1 14 0 0.308744 2 29 0 0.833273 3 6 0 0.109681 4 25 1 0.724122 5 18 1 0.459588 6 4 0 0.081961 7 18 0 0.459588 8 12 0 0.244530 9 22 1 0.618220 10 6 0 0.109681 11 30 1 0.854458 12 11 0 0.216022 13 30 1 0.854458 14 5 0 0.094919 15 20 1 0.539915 16 13 0 0.275479 17 9 0 0.166450 18 32 1 0.890123 19 24 0 0.690829 20 13 1 0.275479 21 19 0 0.499750 22 4 0 0.081961 23 28 1 0.809690 24 22 1 0.618220 25 8 1 0.145294 MTB > print c1 c2 c12 c22 c8 ROW C1 C2 C12 C22 C8 1 14 0 0.349199 0.340294 0.308744 2 29 0 0.822122 0.830373 0.833273 3 6 0 0.096973 0.078918 0.109681 4 25 1 0.696009 0.699685 0.724122 5 18 1 0.475312 0.470981 0.459588 6 4 0 0.033916 0.013574 0.081961 7 18 0 0.475312 0.470981 0.459588 8 12 0 0.286142 0.274950 0.244530 9 22 1 0.601425 0.601669 0.618220 10 6 0 0.096973 0.078918 0.109681 11 30 1 0.853651 0.863045 0.854458 12 11 0 0.254614 0.242278 0.216022 13 30 1 0.853651 0.863045 0.854458 14 5 0 0.065444 0.046246 0.094919 15 20 1 0.538368 0.536325 0.539915 16 13 0 0.317670 0.307622 0.275479 17 9 0 0.191557 0.176934 0.166450 18 32 1 0.916707 0.928388 0.890123 19 24 0 0.664481 0.667013 0.690829 20 13 1 0.317670 0.307622 0.275479 21 19 0 0.506840 0.503653 0.499750 22 4 0 0.033916 0.013574 0.081961 23 28 1 0.790594 0.797701 0.809690 24 22 1 0.601425 0.601669 0.618220 25 8 1 0.160029 0.144262 0.145294 MTB > mplot c12 c1, c8 c1, c2 c1 1.05+ - C C C C 2 C C 2 C - 2 - B 2 4 - A 0.70+ B 2 - 4 A - 2 - 4 2 - 0.35+ 2 A - A A 2 B - A B B - 2 2 B - 2 2 2 0.00+ 4 C 2 C C C C C C C C C - --+---------+---------+---------+---------+---------+---- 5.0 10.0 15.0 20.0 25.0 30.0 A = C12 vs. C1 C = C2 vs. C1 B = C8 vs. C1 MTB > plot c8 c1 - 0.90+ * - * 2 C8 - * - * * - 0.60+ 2 - * - 2 * - - 0.30+ 2 * - * * - * - * 2 * - 2 0.00+ --+---------+---------+---------+---------+---------+----C1 5.0 10.0 15.0 20.0 25.0 30.0 ================================================ NEW BLOG ver 11 MTB > Read "G:\DRR95\95ED257\WEBARCH\PROGRAM.DAT" c1-c2. Entering data from file: G:\DRR95\95ED257\WEBARCH\PROGRAM.DAT 25 rows read. MTB > describe c1-c2 Descriptive Statistics Variable N Mean Median Tr Mean StDev SE Mean C1 25 16.88 18.00 16.78 9.08 1.82 C2 25 0.440 0.000 0.435 0.507 0.101 Variable Min Max Q1 Q3 C1 4.00 32.00 8.50 24.50 C2 0.000 1.000 0.000 1.000 MTB > blogist c2 = c1 Binary Logistic Regression Response Information Variable Value Count C2 1 11 0 14 Total 25 Logistic Regression Table Odds 95% CI Predictor Coef StDev Z P Ratio Lower Upper Constant -3.060 1.259 -2.43 0.015 C1 0.16149 0.06498 2.49 0.013 1.18 1.03 1.33 Log-Likelihood = -12.712 Test that all slopes are zero: G = 8.872, DF = 1, P-Value = 0.003 Goodness-of-Fit Tests Method Chi-Square DF P Pearson 19.623 17 0.294 Deviance 19.879 17 0.280 Hosmer-Lemeshow 5.946 8 0.653 Measures of Association: (Between the Response Variable and Predicted Probabilities) Pairs Number Percent Summary Measures Concordant 127 82.5% Somers D 0.66 Discordant 25 16.2% Goodman-Kruskal Gamma 0.67 Ties 2 1.3% Kendalls Tau-a 0.34 Total 154 100.0% MTB > blogist c2 = c1; SUBC> brief 3. Binary Logistic Regression Response Information Variable Value Count C2 1 11 0 14 Total 25 Factor Information Logistic Regression Table Odds 95% CI Predictor Coef StDev Z P Ratio Lower Upper Constant -3.060 1.259 -2.43 0.015 C1 0.16149 0.06498 2.49 0.013 1.18 1.03 1.33 Log-Likelihood = -12.712 Test that all slopes are zero: G = 8.872, DF = 1, P-Value = 0.003 Goodness-of-Fit Tests Method Chi-Square DF P Pearson 19.623 17 0.294 Deviance 19.879 17 0.280 Hosmer-Lemeshow 5.946 8 0.653 Brown: General Alternative 0.134 2 0.935 Symmetric Alternative 0.086 1 0.769 Measures of Association: (Between the Response Variable and Predicted Probabilities) Pairs Number Percent Summary Measures Concordant 127 82.5% Somers D 0.66 Discordant 25 16.2% Goodman-Kruskal Gamma 0.67 Ties 2 1.3% Kendalls Tau-a 0.34 Total 154 100.0% MTB >