The International Rescue Committee PRO-Jeunes Program Impact Evaluation Surveys, 2019-2023

Baseline and Midline

General Procedure :
To be eligible for the PRO-Jeunes program, applicants must be out-of-school or have dropped out of school, and show proof of identity. To attract participants, the PRO-Jeunes program has been advertised in a variety of ways. Recruitment takes place in three stages. First, young people interested in the program register and express their interest. Secondly, the program team reviews all applications and calls eligible candidates to provide further information about the program and confirm their interest in taking part. Thirdly, candidates who have completed all the previous steps are invited to a registration meeting. Participants at the registration meeting who have enrolled in a training course constitute the list from which we draw our sample.
From this list, we randomly sample, stratifying by gender, living area and choice of training.

Stratification variables :
Baseline 1: gend_r, zone, id_choix,
Baseline 2: gendc, zone, id_choix

Specifics for each Baseline :

Baseline 1: The aim was to conduct individual interviews with out-of-school young people aged between fifteen (15) and twenty-four (24) across the ten communes of the Abidjan district and the commune of Grand Bassam. Despite the difficulties encountered in the field, we interviewed 1865 young people out of the 2200 planned, i.e. a completion rate of 84.77%. However, we also used the reserve list and the complementary list, made up of 255 and 181 respectively, to carry out the various replacements throughout the collection period.

Baseline 2:
1125 interviews were expected from each of the A, B and C questionnaires. In view of the refusals, the researchers decided that the refusals should be replaced. Thus, these replacements were to carry out interviews A, B and C in succession. We received two replacement lists. The initial sample contained 1121 people, the first replacement list 227 and a second list of 184. The difficulty of the replacements lay in respecting the strata of each respondent. In other words, for each person to be replaced, we had to find an individual of the same sex, living in the same area and having chosen the same training. It has to be said that the strata were not of the same size either. The number of replacements to be made was often greater than the number of respondents remaining and available. The collection team explained the situation to the researchers, who gave them permission to make replacements outside the strata, while minimizing the gaps. This enabled us to achieve a sample size of 1,125 for surveys A, B and C, with small discrepancies in some strata.

For more information, please refer to the Survey Design and Methodology - Socioeconomic Survey of Refugees in Ethiopia (SESRE) available under Downloads.