Table 1. Logistic regression results for the Cleveland Clinic data.
Variable | Original | modgo | sbgcop-modgo | SimMultiCorrData |
---|---|---|---|---|
Exercise-induced angina (yes) | 4.43 | 4.45 (2.38–8.17) | 10.16 (5.52–19.26) | 4.55 (2.78–8.04) |
Age (years) | 1.02 | 1.02 (0.98–1.06) | 1.05 (1.01–1.08) | 1.02 (0.99–1.06) |
Max heart rate (bpm) | 0.98 | 0.97 (0.95–0.99) | 1.00 (0.99–1.02) | 0.97 (0.96–0.98) |
ST depression (mm) | 1.92 | 2.07 (1.52–2.86) | 1.60 (1.26–2.23) | 2.18 (1.69–2.86) |
Displayed are odds ratios for the original data (column 2). Three simulation methods were employed, and the medians of the odds ratios and empirical 2.5 and 97.5% quantiles of the odds ratios (in parenthesis) are displayed from 500 simulated datasets per simulation method. Column 3 shows results when simulations were done with modgo, column 4 when modgo was used with the correlation matrix estimated by sbgcop, and column 5 when simulations were done with SimMultiCorrData.