Prev: Out of memory Error for Unstructured correlation structure in GEE
Next: Please help! ----SAS IML for Newton Iteration, Weibull distribution...
From: barro on 14 Apr 2010 14:01
Hello,

I am trying to estimate a difference in difference regression in SAS
but am not having much luck. I was hoping someone could advise me on
how to code for the variables I am using (I know how to interpret but
not run:( )

My study is on the effects of welfare reform on labor force
participation and hours worked by single mothers w/ <h.s. diploma. I
am using two control groups one married mothers w/ less than h.s.
diploma, one single mothers w/ greater than h.s. diploma. I have two
sets of individual data downloaded from the March CPS supplement and
the following variables:
PreTANF coded 0 or 1, 1 = before policy
PostTANF coded 0 or 1, 1 = after policy
Single coded 0 or 1, 1 single
H.S., coded 0, 1, 1 = h.s.
college, coded 0, 1, 1=some college
Race, coded 0,1, 1 = white
Edu - class variable for educational attainment 1 = elementary, 2 =
middle, 3 = high school, no diploma, 4 = high school diploma, 5 = some
college...etc.
Numb.Children - number of children

I am trying to regress Labor force participation (0, 1 variable w/ 1
being in labor force) on PostTANF and do a difference in difference
regression. I am confused on how to do this for many reasons. First, I
am not sure how to model. Second, I don't know what regression to use
(logistic?) w/ a 0,1 variable.

Any advice much appreciated!!!!!

Thanks.







 | 
Pages: 1
Prev: Out of memory Error for Unstructured correlation structure in GEE
Next: Please help! ----SAS IML for Newton Iteration, Weibull distribution...