Skip to main content

Table 2 Number of samples among 1000 leading to outliers in GMM estimation

From: Instrumental variable analysis in the context of dichotomous outcome and exposure with a numerical experiment in pharmacoepidemiology

n

Level

Weak

Mod

Strong

30000

    
 

High

155

61

31

 

Med

78

65

43

 

Low

41

149

65

20000

    
 

High

149

70

64

 

Med

64

72

43

 

Low

25

124

−

10000

    
 

High

95

58

37

 

Med

−

78

55

 

Low

−

110

44

  1. Legend: Low, Med (Medium), High denote level of confounding whereas Weak, Mod (Moderate), Strong stand for instrument strength. The number n with values 10000, 20000 and 30000 stands for the sample size
Back to article page

BMC Medical Research Methodology

ISSN: 1471-2288

Contact us