Castro-Pearson, Sandra2023-09-192023-09-192023-01https://hdl.handle.net/11299/257019University of Minnesota Ph.D. dissertation. January 2023. Major: Biostatistics. Advisors: Xianghua Luo, Chap Le. 1 computer file (PDF); x, 97 pages.Randomized controlled trials are the most reliable form of evidence to drive medical practice and health policy. Yet, challenges remain in terms of understanding possible errors during protocol implementation and interpreting results of complex outcomes such as censored time-to-event outcomes with and without competing risks. In this dissertation, we study receiver operating characteristic (ROC) curves and use ROC concepts to create graphical methods that address these challenges.enClinical TrialsCompeting riskKaplan-MeierROCSurvival analysisThesis or Dissertation