Raising the Bar: Impacts and Implementation of the New Heights Program for Expectant and Parenting Teens in Washington, dc

This page has been left blank for double-sided copying. 

---

APPENDIX B 
 
ADMINISTRATIVE DATA PREPARATION 

---

This page has been left blank for double-sided copying. 

---

APPENDIX B: ADMINISTRATIVE DATA PREPARATION
 
In this appendix, we provide an overview of the administrative data sources used to estimate 
the impacts of New Heights expansion, including a description of each data source, details on 
how we linked the data, and information on how we accounted for changes over time in 
administrative record keeping that affected some outcome variables. In addition, we describe 
how we formed our primary analytic sample and provide descriptive statistics for that sample.
Data sources 
As described in Chapter III, the estimation of New Heights impacts relied on data from three 
administrative sources: (1) DC Public Schools (DCPS), (2) DC Department of Health (DC 
DOH), and (3) DC Department of Human Services (DC DHS). This section describes these data 
sources, our method of linking them, and changes to the way DC measured outcome variables 
over time.
Description of data sources 
DCPS provided two data files containing student records. One file contained student 
demographic and outcome information; we used this to identify teens attending any DCPS high 
school from the fall semester of the 2007–2008 school year through the spring semester of the 
2014–2015 school year. The second file from DCPS contained school admission and withdrawal 
records. We used this file to identify teens who were enrolled in a study school.
We used DC DOH vital statistics records to identify teens who gave birth in the District of 
Columbia. These data include birth date, mother’s name, and mother’s address as reported on the 
birth certificate. For the purpose of our analysis, DC DOH limited the data file to the 12,539 
children born to mothers ages 14–19 in a D.C. hospital from January 1, 2003, through December 
31, 2014. We included DC DOH data from four and a half years before the earliest observed 
DCPS record so that parenting females who gave birth before the earliest observed records in the 
DCPS data can be correctly identified as parents. Although it is possible for parenting females in 
DCPS to have given birth outside of a D.C. hospital—and for the birth to therefore be missing 
from DC DOC data—New Heights coordinators report that this would be a rare circumstance, as 
most parenting females received D.C. Medicaid and attended prenatal providers with admitting 
permission for D.C. hospitals only.
The DC DHS maintains the New Heights participation database on behalf of the program. 
These data include all of the 1,371 New Heights participants from the fall semester of the 2011–
2012 school year through the spring semester of the 2014–2015 school year. We used these data 
to identify New Heights participants to estimate the impact of New Heights on participants. 
These data also include information used in the implementation analysis (see Appendix A).
The combined administrative data set enabled us to (1) identify parenting teens from DOH 
birth records, (2) identify which of those parenting teens were enrolled in a DCPS study school 
using DCPS records, (3) measure outcomes expected to be affected by New Heights, and (4) 
determine whether parenting teens in study schools participated in the New Heights program. 
The next section describes our method of constructing the combined administrative data set from 
the DCPS, DC DOH, and DC DHS data.
B.3 

---

APPENDIX B: ADMINISTRATIVE DATA PREPARATION