Skip to main content
. Author manuscript; available in PMC: 2023 May 10.
Published in final edited form as: Stat Med. 2022 Feb 8;41(10):1751–1766. doi: 10.1002/sim.9326

Figure 5:

Figure 5:

Validity of the two proposed models in the disinvestment example if they were both misspecified. Throughout, τt = 0.28 and σt = 1.03 were fixed, based on their average fitted values from the two models. We assumed there were K=177 observations per cluster per time period, which was the average K across the trial.