Summarised simulation results - regression coefficients

Summarised simulation results - regression coefficients

regr_results

Format

var

Variable for which a regression coefficient is being estimated, for example "X.1" corresponds to the cause-specific effect of X on event 1 (Relapse)

m

Number of imputed datasets

analy

Method indicator, please refer to manuscript for abbreviations used

true

True data-generating value for regression coefficient

scen_num

Number label for scenario

n

Sample size in simulated datasets

est

Estimated regression coefficient

se

Estimate model standard error (SE) for regression coefficient

se_mcse

Monte carlo standard error for model SE

emp_se

Empirical SE

cover

Coverage

bias

Bias

rmse

Coverage

rmse_mcse

Coverage

warns

Number of warnings across all simulations of a particular scenario, for a coefficient. Generally corresponds to number of smcfcs rejection sampling failures.

bias_mcse

Monte carlo error of bias

cover_mcse

Monte carlo error of coverage

prop_miss

Proportion of missing values in X

beta1

Data-generating ffect of X on event 1

miss_mech

Missingness mechanism

X_level

Measurement level of X

rho

Correlation between X and Z

eta1

Strength of missingness mechanism for non-MCAR scenarios

haz_shape

Shapes of baseline hazard for competing events, either "similar" or "different"