Skip to main content

Table 4 Percentages of selecting \(\mathcal {M}_{1}\) according to the BIC on N=500 datasets, given t v =(0,1,2,4,6,8,10,12) and \(\sigma _{0}^{2}=1.5\)

From: Item response models for the longitudinal analysis of health-related quality of life in cancer clinical trials

Parameter

Scenarios

 

Values

AM using δ ne

CM using δ fa

CM using δ ne

AM using δ fa

\(\sigma _{1}^{2}\)

β 1

LMM

AM

CM

LMM

AM

CM

LMM

AM

CM

LMM

AM

CM

0.2

−0.3

0

0

0

0

0

0

0

0

0

0

0

0

0.2

0.3

0

0

0

0

0

0

0

0

0

0

0

0

0

−0.5

97.7

99.3

56.49

100

94.6

93.0

100

61.3

95.7

100

99.7

89.5

0

−0.3

99.0

100

33.0

100

88.6

93.3

100

36.3

94.9

100

100

83.3

0

−0.2

100

99.6

49.3

100

94.6

93.8

100

71.7

95.8

100

99.6

79.0

0

−0.1

98.7

95.7

94.8

100

98.7

89.6

100

99.0

90.4

100

100

88.1

0

0.0

95.6

100

94.6

99.0

99.7

91.8

99.0

99.7

89.7

97.0

99.7

94.4

0

0.1

83.0

100

94.8

93.3

100

92.6

97.0

100

90.9

87.3

100

94.7

0

0.3

98.3

99.6

90.6

100

99.6

89.1

100

100

93.7

100

99.6

93.8

0

0.5

100

100

94.3

100

99.3

94.7

100

100

97.6

100

100

97.2

  1. The (adjacent,logistic, Z 1,U a ) a=1,2 models and the (cumulative,logistic, Z 1,U a ) a=1,2 models are denoted respectively by AM and CM. For the random component, U 1 if \(\sigma _{1}^{2}=0\) and U 2 if \(\sigma _{1}^{2}>0\)