Table 16 Mantel–Haenszel bounds sensitivity analysis for IPV (N = 59,093).

From: The association between access to key household resources and violence against women

Variables

Γ

MH statistic

Sig. level

Overestimation of TE

Underestimation of TE

Overestimation of TE

Underestimation of TE

Drinking water

1

18.332

18.332

0.000

0.000

1.5

39.489

2.426

0.000

0.008

2

55.024

17.169

0.000

0.000

2.5

67.548

28.699

0.000

0.000

3

78.192

38.220

0.000

0.000

3.5

87.547

46.364

0.000

0.000

4

95.961

53.503

0.000

0.000

4.5

103.653

59.874

0.000

0.000

5

110.775

65.639

0.000

0.000

Toilet facility

1

34.175

34.175

0.000

0.000

1.5

56.154

12.866

0.000

0.000

2

72.460

2.070

0.000

0.019

2.5

85.699

13.676

0.000

0.000

3

97.008

23.207

0.000

0.000

3.5

106.986

31.324

0.000

0.000

4

115.988

38.416

0.000

0.000

4.5

124.239

44.728

0.000

0.000

5

131.895

50.429

0.000

0.000

  1. Γ is the odds of differential assignment due to unobserved factors.
  2. TE treatment effects.