Implementing the MI-Wilson Confidence Interval


[Up] [Top]

Documentation for package ‘MIWilson’ version 1.0.0

Help Pages

Bm Calculate between-imputation variance of the response mean \frac{sum (\hat{Q}_l-\bar{Q})}{m-1}
dof Calculate degrees of freedom used in calculating confidence intervals of t-distributed proportion point estimate \bar{Q}_m
mi_wald Calculates the specified Wald CI of a binomial proportion variable, given imputed data sets.
mi_wald_phat Calculates the MI-Wald interval if given a vector of observed binomial proportions (one for each imputed data frame)
mi_wilson Calculates the specified Wilson CI of a binomial proportion variable, given imputed data sets.
mi_wilson_phat Calculates the MI-Wilson interval if given a vector of observed binomial proportions (one for each imputed data frame)
Qbar Calculate Qbar (average response over MICE datasets)
Qhats Calculate Qhats (means of response for each imputed dataset)
Rm Helper function for getting rm, a key component for calculating degrees of freedom and the wilson CI directly
Tm Estimate variance of proportion point estimate \bar{Q}_m
Ubar Calculate Ubar (average response variance over MICE datasets)
Uhats Calculate Uhats (variance for each imputed dataset)