Skip to main content

Table 2 OLS regression (HIV testing)

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

 CoefbSECI (lower)CI (higher)Coverage probability (%)MSE
Women
 Original dataa
  Spatial lag1.1590.2260.7161.603  
  Wealth0.0030.003− 0.0030.010  
  Education0.0080.007− 0.0070.022  
 5000 simulation
  Wealth0.0080.009− 0.0100.02694.50.0001
  Education0.0110.018− 0.0250.04697.80.0003
 10,000 simulation
  Wealth0.0080.009− 0.0100.02695.00.0001
  Education0.0110.018− 0.0240.04697.90.0003
Men
 Original dataa
  Spatial lag1.3370.1711.0011.672  
  Wealth− 0.0110.002− 0.016− 0.007  
  Education0.0180.0050.0090.027  
  _cons− 0.2700.144− 0.5520.013  
 5000 simulation
  Wealth− 0.0100.009− 0.0290.00995.10.0003
  Education0.0190.019− 0.0190.05696.60.0005
 10,000 simulation
  Wealth− 0.0100.009− 0.0280.00895.60.0001
  Education0.0190.019− 0.0190.05697.00.0004
  1. MSE mean squared error
  2. aRow normalised
  3. bThis was estimated using spmat and spreg package in Stata