Skip to main content
Fig. 3 | Diagnostic and Prognostic Research

Fig. 3

From: Reclassification calibration test for censored survival data: performance and comparison to goodness-of-fit criteria

Fig. 3

Power when an important predictor is omitted for 0, 25, and 50% censoring rates. The RC statistic was calculated for the full (y~ x1 + x2) and reduced (y~ x1) models. Data were simulated according to the full model, but the reduced model was used to estimate the expected proportions in the RC statistic formula. Left panel—decreasing baseline hazard, right panel—increasing baseline hazard. N = 1000 (top row), 5000 (middle row), 10,000 (bottom row), p = 0.1, collapse when evg < 5

Back to article page