https://ckitsignal.com/index.p....hp/a-sensible-exampl
Simulation scenarios varied by size of subgroup, energy of association of subgroup with exposure, power of connection of subgroup with outcome (simulated success), and outcome occurrence. Results indicated that subsetting the data because of the subgrouping variable, to estimate the propensity score and threat ratio, gets the tiniest bias, far exceeding any penalty in precision. Furthermore, weighting techniques spend a heavier price in prejudice than do matching techniques whenev
Mi piace
Commento
Condividi
