Soohyung Lee and Azeem M. Shaikh, "Multiple Testing and Heterogeneous
Treatment Effects: Re-evaluating the Effect of PROGRESA on School
Enrollment", Journal of Applied Econometrics, Vol. 29, No. 4, 2014,
pp. 612-626.

The file named ls-data.csv contains the dataset. It is zipped in the
file ls-files.zip. Unix/Linux users should use "unzip -a".

The data were extracted from the original data provided by the Mexican
government available at

  http://evaluacion.oportunidades.gob.mx:8010/en/index.php

These data are saved in an ASCII file in a comma-separated format. The
file includes 11 variables in the following order.

Column 1: wave of survey

Column 2: village identifier

Column 3: observation number within a village in a survey

Column 4: village size

Column 5: whether the corresponding village is subject to PROGRESA

Column 6: individual identifier

Column 7: 1 if child is female

Column 8: age

Column 9: highest grade completed

Column 10: 1 if child is eligible to get a support from PROGRESA if
           he/she lives in a village subject to PROGRESA

Column 11: 1 if child is enrolled in a school.

The file called "leeandshaikh_mht_code.m" is the matlab code that
conducts the multiple hypothesis testing procedure. The code produces
results reported in the following two tables. It is also zipped in
ls-files.zip.

AtableDI: Column 1 reports the mean difference in school enrollment
rate between treated and untreated in a subgroup. Columns 2 to 5 are
reported in columns 1 to 4 of Table 2.

AtableDID: Column 1 reports the DID in school enrollment rate. Columns
2 to 5 are reported in columns 5 to 8 of Table 2.

We include some comments in the code for further explanation.