Skip to main content

Table 2 Summary of empirical estimates of log-OR and relative bias (%), power (%) and MC standard error, coverage (%), model standard error and empirical standard error derived from χ2 method and GLMM test with compound symmetric variance-covariance matrix from 1000 simulations with a total sample size N = 398

From: Power difference in a χ2 test vs generalized linear mixed model in the presence of missing data – a simulation study

 

pi1 = 0.1, ρ = 0.7

Log-odds (% bias)

Power (%)c

Coverage (%)c

Model SEc

Empirical SEc

Complete

χ2 a

0.792 (0)

78.2

95.8

0.301

0.295

 GLMMb

0.944 (0)

84.7

98.1

0.335

0.292

MCAR

Equal drop out (30% drop out in each arm)

χ2

0.798 (0.7)

62.3

95.0

0.362

0.367

  GLMM

0.953 (0.9)

73.5

98.4

0.384

0.328

Unequal dropouts 1 (20% drop out in control, 40% drop out in treatment)

χ2

0.796 (0.5)

64.6

95.3

0.357

0.354

  GLMM

0.961 (1.8)

74.9

98.3

0.381

0.326

Unequal dropouts 2 (40% drop out in control, 20% drop out in treatment)

χ2

0.803 (1.4)

60.2

95.4

0.373

0.372

  GLMM

0.946 (0.2)

72.2

98.0

0.390

0.329

MAR1 (observations more likely to be deleted if prior observation was positive)

Equal drop out (30% drop out in each arm)

χ2

1.032 (30.2)

73.8

93.3

0.424

0.427

  GLMM

1.075 (13.9)

83.3

97.5

0.400

0.341

Unequal dropouts 1 (20% drop out in control, 40% drop out in treatment)

χ2

0.842 (6.3)

58.1

95.3

0.402

0.406

  GLMM

1.016 (7.6)

79.1

98.0

0.392

0.336

Unequal dropouts 2 (40% drop out in control, 20% drop out in treatment)

χ2

1.194 (50.3)

83.1

89.7

0.455

0.463

  GLMM

1.117 (18.3)

85.0

97.2

0.408

0.346

MAR2 (observations less likely to be deleted if prior observation was positive)

Equal drop out (30% drop out in each arm)

χ2

0.690 (−12.9)

55.6

93.7

0.339

0.336

  GLMM

0.879 (−7.0)

65.3

98.2

0.380

0.325

Unequal dropouts 1 (20% drop out in control, 40% drop out in treatment)

χ2

0.826 (4.3)

69.7

95.8

0.340

0.337

  GLMM

0.935 (−1.0)

72.3

98.4

0.380

0.326

Unequal dropouts 2 (40% drop out in control, 20% drop out in treatment)

χ2

0.577 (−27.2)

37.3

89.1

0.342

0.346

  GLMM

0.831 (−12.0)

58.4

97.5

0.382

0.328

  1. aIndependent two-sample χ2 test for test of treatment effect at time point 3
  2. bGeneralized linear mixed model with compound symmetric variance-covariance matrix, contrast used to estimate treatment effect at time point 3
  3. cRange of Monte Carlo SE: Power (0.011–0.016); Coverage (0.004–0.010); Model SE (0.0002–0.003); Empirical SE (0.007–0.010)
Back to article page

BMC Medical Research Methodology

ISSN: 1471-2288

Contact us