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

Analysis approach.
Qualitative analysis of the interviews, focus groups with youth, and 
observation data involved an iterative process using thematic analysis and triangulation of data 
sources (Patton 2002
21
; Ritchie and Spencer 1994
22
). Trained staff used a qualitative analysis 
software package, Atlas.ti, to facilitate organizing and synthesizing the qualitative data. First, we 
developed a coding scheme for the study, organized according to key research questions. Within 
each question, we defined codes for key themes and subtopics. Then, we applied the codes to 
passages in the interview, observation, and focus group notes. To ensure accurate and consistent 
coding, two research assistants first independently coded site visit data. Then two members of 
the site visit team reviewed the coded documents and reconciled any differences in coding. To 
address the research questions, we used the software to retrieve relevant passages and then 
examined the patterns of responses across respondents and identified emergent themes. 
New Heights program database 
We used administrative data from the DC Department of Human Services’ (DC DHS) New 
Heights database to analyze (1) the number of workshops offered by New Heights, by topic, in 
school year 2014–2015 and (2) the average workshop attendance rate for New Heights 
participants in school year 2014–2015. We limited both analyses to female New Heights 
participants in 9 of the 11 study schools. Two schools (Roosevelt STAY and Ballou STAY) were 
excluded from both workshop analyses because they were also excluded from the impact 
analyses. 
Workshops offered:
To calculate the number of workshops offered by topic, we used data 
from the DHS workshops table, which reports each workshop a school offers, including its topic, 
subtopic, date offered, and number of attendees. Workshops were tallied by topic and subtopic, 
and those with zero attendees were excluded from the analysis.