Skip to main content

Table 3 Power of Methods for Missing Data Percentages (10% and 40%) and Mechanisms

From: A comparison of missing data methods for hypothesis tests of the treatment effect in substance abuse clinical trials: a Monte-Carlo simulation study

Method

Missing Rate

MCAR

MAR

MAR/MNAR

MNAR

SSSa

10%

0.8370

0.9355

0.9225

0.9355

 

40%

0.5705

0.6125

0.5750

0.5600

Modified SSSb (Dawson Weight)

10%

0.8165

0.9000

0.8810

0.8970

 

40%

0.4445

0.3755

0.3660

0.3000

Modified SSSb (Population Weight)

10%

0.8240

0.9075

0.8910

0.9075

 

40%

0.4990

0.4150

0.4105

0.3850

Modified SSSb (SSTime Weight)

10%

0.8220

0.8675

0.8820

0.8990

 

40%

0.4770

0.3075

0.3795

0.3780

Fisherc

10%

0.7655

0.6940

0.6710

0.6855

 

40%

0.3240

0.1505

0.1460

0.1195

Weighted Z-Transformd

10%

0.8265

0.9080

0.8965

0.9085

 

40%

0.4990

0.4440

0.4355

0.4130

  1. *Power, r = .6, n = 100, simulations = 2000
  2. *a, b, c, d indicate the methods of analysis in the text