6 Simulation study: impact of misspecification in survival models with shared frailty terms

In this Chapter, I will present the second simulation study I set up and run during my first year. It investigates the impact of model misspecification in survival models with shared frailty terms, and part of this work was presented in oral form at the 2017 Statistical Analysis of Multi-Outcome Data (SAM) Conference and at the 38^th Annual Conference of the International Society for Clinical Biostatistics (more details in Chapter 10 and slides available in Appendix C.2). I am also currently writing up this project into a manuscript for submission to a journal; a current draft is attached as well in Appendix D.

This Chapter is arranged as follows. First, I introduce the aim of the simulation study in greater detail in Section 6.1. Then, in Section 6.2 I will introduce the data-generating mechanisms, in Section 6.3 I will describe the different models I included in the comparison, in Section 6.4 I will define the estimands of interest, in Section 6.5 I will present the performance measures used to compare the different models, in Section 6.6 I will present some results, and finally I will conclude the Chapter in Section 6.7.