Skip to main content

Table 6 The effect of religious belief on the health of the elderly (CESD_d)

From: The effect of religious belief on Chinese elderly health

Matching methodMatching parameterCommon support sample sizeCESD_dS.E.
TreatedControlsATT
One-to-one matchingk = 1699−0.054−0.0660.0110.040
K-nearest neighbor matchingk = 104016−0.054−0.0590.0040.032
Radius matchingδ=0.014015−0.054−0.0570.002*0.030
Kernel matchingk: epan
bw: 0.06
4016−0.054−0.052− 0.003*0.029
  1. Notes: *p < 0.1. CESD_d is the improvement of self-report depression scale between 2012 and 2016. ATT is the average treatment effect on the treated. S.E. is standard error that estimated by bootstrap (with 500 replications). K is the number of neighborhoods for matching; δ is the matching radius; Epan is the kernel function; Bw is the bandwidth