The rest are binary categorical data.įinally, the last variable named Earnings in 1978 is the total revenue of the participant in the year 1978. No degree: 1 if the participant did not receive any degree, 0 otherwise Įarnings in 1974: the total revenue of the participant in 1974 expressed in 1978 dollars Įarnings in 1975: the total revenue of the participant in 1975 expressed in 1978 dollars Īge, years of education, earnings in 19 are quantitative data. Married: 1 if the participant is married, 0 otherwise Hispanic-American: 1 if the participant is of Hispanic-American origin, 0 otherwise Years of education: the number of years of education for the participant Īfro-American: 1 if the participant is of Afro-American origin, 0 otherwise The suspected confounding variables are the 8 variables below:Īge: the age of the participant in years We believe that the probability for one participant to participate in the job training program is not random and that it can be explained to some extent with confounding variables. The first one is binary variable entitled Participation in job training that indicates if the participant did participate in the job training program (1) or not (0). There are 10 variables used in the original study. This will be achieved by pairing similar individuals in terms of confounding variables using Propensity Score Matching. Some confounding variables are suspected to introduce a serious bias in the study results so we wish to quantify this effect and build up a subsample of it that reduces this bias beforehand. The original study aimed at studying the effect of the participation in a particular job training program on the earnings of the individuals in 1978.
#Xlstat stastical software full
The full dataset is available at this address. This tutorial uses a random subsample of a dataset originally published in Robert Lalonde (1986) and revisited in Dehejia and Wahba (1999). Dataset for running a propensity score matching This tutorial will help you set up and perform a propensity score matching in Excel using the XLSTAT statistical software.
0 kommentar(er)
