JC: Job Corps data

In causalweight: Estimation Methods for Causal Inference Based on Inverse Probability Weighting

Job Corps data

Description

A dataset from the U.S. Job Corps experimental study with information on the participation of disadvantaged youths in (academic and vocational) training in the first and second year after program assignment.

Usage

JC

Format

A data frame with 9240 rows and 46 variables:

assignment

1=randomly assigned to Job Corps, 0=randomized out of Job Corps

female

1=female, 0=male

age

age in years at assignment

white

1=white, 0=non-white

black

1=black, 0=non-black

hispanic

1=hispanic, 0=non-hispanic

educ

years of education at assignment

educmis

1=education missing at assignment

geddegree

1=has a GED degree at assignment

hsdegree

1=has a high school degree at assignment

english

1=English mother tongue

cohabmarried

1=cohabiting or married at assignment

haschild

1=has at least one child, 0=no children at assignment

everwkd

1=has ever worked at assignment, 0=has never worked at assignment

mwearn

average weekly gross earnings at assignment

hhsize

household size at assignment

hhsizemis

1=household size missing

educmum

mother's years of education at assignment

educmummis

1=mother's years of education missing

educdad

father's years of education at assignment

educdadmis

1=father's years of education missing

welfarechild

welfare receipt during childhood in categories from 1 to 4 (measured at assignment)

welfarechildmis

1=missing welfare receipt during childhood

health

general health at assignment from 1 (excellent) to 4 (poor)

healthmis

1=missing health at assignment

smoke

extent of smoking at assignment in categories from 1 to 4

smokemis

1=extent of smoking missing

alcohol

extent of alcohol consumption at assignment in categories from 1 to 4

alcoholmis

1=extent of alcohol consumption missing

everwkdy1

1=has ever worked one year after assignment, 0=has never worked one year after assignment

earnq4

weekly earnings in fourth quarter after assignment

earnq4mis

1=missing weekly earnings in fourth quarter after assignment

pworky1

proportion of weeks employed in first year after assignment

pworky1mis

1=missing proportion of weeks employed in first year after assignment

health12

general health 12 months after assignment from 1 (excellent) to 4 (poor)

health12mis

1=missing general health 12 months after assignment

trainy1

1=enrolled in education and/or vocational training in the first year after assignment, 0=no education or training in the first year after assignment

trainy2

1=enrolled in education and/or vocational training in the second year after assignment, 0=no education or training in the second year after assignment

pworky2

proportion of weeks employed in second year after assignment

pworky3

proportion of weeks employed in third year after assignment

pworky4

proportion of weeks employed in fourth year after assignment

earny2

weekly earnings in second year after assignment

earny3

weekly earnings in third year after assignment

earny4

weekly earnings in fourth year after assignment

health30

general health 30 months after assignment from 1 (excellent) to 4 (poor)

health48

general health 48 months after assignment from 1 (excellent) to 4 (poor)

References

Schochet, P. Z., Burghardt, J., Glazerman, S. (2001): "National Job Corps study: The impacts of Job Corps on participants' employment and related outcomes", Mathematica Policy Research, Washington, DC.

Examples

## Not run: 
data(JC)
# Dynamic treatment effect evaluation of training in 1st and 2nd year
# define covariates at assignment (x0) and after one year (x1)
x0=JC[,2:29]; x1=JC[,30:36]
# define treatment (training) in first year (d1) and second year (d2)
d1=JC[,37]; d2=JC[,38]
# define outcome (weekly earnings in fourth year after assignment)
y2=JC[,44]
# assess dynamic treatment effects (training in 1st+2nd year vs. no training)
output=dyntreatDML(y2=y2, d1=d1, d2=d2, x0=x0, x1=x1)
cat("dynamic ATE: ",round(c(output$effect),3),", standard error: ",
    round(c(output$se),3), ", p-value: ",round(c(output$pval),3))
## End(Not run)

causalweight documentation built on May 4, 2023, 5:10 p.m.