A comparison of different measures of the proportion of explained variance in multiply imputed data sets