Experimental Design
The proposed randomized control trial is composed of 300 vulnerable young adults, who will be randomly sorted via lottery during December 2018 and January 2019 into three groups: a control group, a treatment group receiving vocational education (Treatment 1), and a treatment group receiving vocational education, psychosocial support and citizen empowerment training - our life skills training component (Treatment 2). Each group will be composed of 100 young adults.
We will undertake three overall data collections: a baseline data collection consisting of the details provided under “outcomes” above - prior to the training at the beginning of 2019, a mid-line survey upon completion of training after one year (end of 2019, beginning of 2020), and a final data collection in November 2021 to estimate mid-term effects of the program. Thus, the study will compare both the two treatment groups, and the two treatment groups separately to the control group.
During the baseline data collection, participants take part in four surveys:
- The sign-up survey collects details on wages, hours worked, formal and informal employment, administrative details, and the Wartegg test results;
- The psychological interview conducted by the in-house therapist collects information on life expectations, family situation and support, and a subjective evaluation of participants’ leadership style, conflict resolution style, interpersonal relationships, verbal fluidity, self-confidence and presentation;
- The household visit collects information on the infrastructure of the household, and details on family members (how many, income levels, and so on).
- The randomization survey consists in the Global Preferences Survey, the PCL-C, and an adaptation of Blattman et al (2017)’s survey
There will be a selection process to reduce attrition. This includes: study of submitted CV’s; conducting personal interviews and family visits; processing of a Wartegg test. While we understand that this limits the external validity of our experiment, this is a necessary step given the extreme vulnerability of our target population. The criteria for entering the program include: living in Bosa or Soacha; classified as vulnerable according to SISBEN (Index 1 or 2); not being able to finance a professional education by own means; demonstrated motivation and abilities to study one year; demonstrated family support.
Because the data is sensitive, we are keeping data mapping individuals to their administrative data separate from the main database. Individuals will be given random codenames, which will be their unique identifiers in the main database.