Skip to main content
. Author manuscript; available in PMC: 2014 Sep 1.
Published in final edited form as: Soc Sci Med. 2013 May 31;92:27–34. doi: 10.1016/j.socscimed.2013.05.018

Table 4.

Does controlling for attrition affect estimates? (Weighted Least Squares)

VARIABLES (1)
Major Depression		 (2)
Minor Depression
Treated × 6 months1 −0 141*** −0.186***
(0.038) (0.065)
Treated × 12 months2 −0.154*** −0.309***
(0.035) (0.053)
Includes Individual Fixed Effects Yes Yes
3 Includes Attrition Weights3 Yes Yes
Observations 1,1 30 1,1 30
R-squared 0.066 0.151

Standard errors in parentheses are clustered at individual level ***p<0.01

**

p<0.05

*

p<0.1

1

This is the difference in difference estimate at 6 months i.e. the change in the dependent variable between Month 0 and month 6 for the treated group relative to the same change for the comparison group.

2

This is the difference in difference estimate at 12 months i.e. the change in the dependent variable between Month 0 and month 12 for the treated group relative to the same change for the comparison group.

3

Attrition weights are inverse probability weights calculated following the procedure outlined in Fitzgerald, Gottshalk and Moffitt (1998)