Skip to main content

Table 4 Sensitivity analysis—OLS (health service use)

From: A bootstrapping approach for generating an inverse distance weight matrix when multiple observations have an identical location in large health surveys

 CoefSECI (lower)CI (higher)  
Women
 Original dataa
  Spatial lag1.6450.1591.3331.956  
  Wealth− 0.0120.004− 0.020− 0.003  
  Education0.0290.0090.0110.048  
  Constant− 0.3820.094− 0.566− 0.198  
 CoefSEBoot CI (lower)Boot CI (higher)Coverage probability (%)MSE
5000 simulation
 Wealth− 0.0070.012− 0.0300.01795.40.000169
 Education0.0240.025− 0.0250.07496.50.000659
10,000 simulation
 Wealth− 0.0070.012− 0.0300.01695.40.000166
 Education0.0250.025− 0.0240.07596.60.000651
 CoefSECI (lower)CI (higher)  
Men
 Original dataa
  Spatial lag− 0.0530.045− 0.1420.036  
  Wealth− 0.0120.003− 0.018− 0.006  
  Education0.0190.0060.0070.032  
  Constant0.7120.0520.6110.813  
 CoefSEBoot CI (lower)Boot CI (higher)Coverage probability (%)MSE
5000 simulation
 Wealth− 0.0140.013− 0.0390.01195.30.000165
 Education0.0190.028− 0.0360.07395.20.000773
10,000 simulation
 Wealth− 0.0140.013− 0.0380.01195.40.000161
 Education0.0180.027− 0.0350.07295.80.000752
  1. aRow normalised