FELLOW TRAINING
Enhancing DATA SCIENCE skills among fellows
Contexte
SEJEN CI’s participation in the Data Analytics Center (D.A.C.) project consisted in providing technical support to the DCDJ (Des Chiffres et des Jeunes) project. With this in mind, SEJEN has set up a training center to ensure capacity building in Data Science for people recruited through the D.C.D.J. project. The aim of this action isb to promote efficient decision-making by organizations (public/private) and populations through the use of reliable data.
Problems
A general observation was made: the people responsible for collecting and compiling data have little or no awareness of data culture, and in particular how to process and use data effectively. Based on this observation, and in line with the orientations of the Data Fellows project, we asked ourselves the following questions:
>How can we build Fellows’ skills in data collection and analysis in just 2 months, to give them the ability to impact the data ecosystem?
>How can Fellows develop the critical thinking skills they need to understand the value of the data they collect or make available, in order to impact the activities of the companies in which they will be assigned during their internships?
>How can we make these Fellows operational?
>How can we motivate young girls to embrace a career in data science, given that according to the UNESCO report “Unlocking the code: educating girls and women in science, technology, engineering and mathematics (STEM)”, in Côte d’Ivoire, only 16% of women are enrolled in quantitative science and engineering training programs?
Solutions
The accelerated Data Science training program targeted two categories of Fellows: Technicals and Staffs.
The Technicals were recruited from graduate students in various relevant fields: Mathematics, Statistics, Economics, Computer Science, etc., to better support the demanding Data Science program.
Training for each cohort lasted 2 months, and covered topics such as Data Visualization, Database Management, Statistics and Machine Learning.
Awareness-raising campaigns were carried out for girls, raising their participation rate in the training to 17% for cohort 1, 30% for cohort 2, and 47% for cohort 3.
These rates, higher than 16%, have contributed to increasing the UNESCO rate of girls in science courses in Côte d’Ivoire.
Methodology
Technical Fellows selection process
To achieve the program’s objectives, Technical Fellows were selected according to a process that can be broken down into 5 stages: Exploration, Recruitment, Pre-Training, Training Deployment and, finally, Follow-up of Fellows during their training period. During the training of each cohort, lessons were learned. These lessons have been incorporated as the process has improved in terms of best practice.
Selection process for Staff Fellows
Staff Fellows were recruited from among employees of PEPFAR’s Implementing Partners (IPs). This category of Fellows had the best backgrounds as well as a good willingness to learn (they were assessed via an online test). The final selection of these Staff Fellows was made after an interview to participate in the Data Science training program.
Course methodology
The methodology applied was participative, practical and based on the following techniques:
– Theoretical courses ;
– Practical work ;
– Questions & Answers ;
– Individual and group work;
– Evaluation before and after each training module.
Results and impact
Fellows’ general level of knowledge significantly increased
Taking cohort 1 as an example, Fellows scored an average of 13.57 out of 30 before the start of each module (general pre-test). This score gives an idea of the fellows’ overall knowledge of data science before starting the training. After receiving the training modules, when subjected to the same test (general post-test), they obtained an average of 22.35 out of 30. This represents a real increase in their overall level of knowledge of +64.65%.
Despite the very demanding and intense level of this accelerated Data Science training program, we were able to achieve above-average success rates, broken down by cohort as follows:
• Cohort 1 : 68,97%
• Cohort 2 : 57,14%
• Cohort 3: 100%
To validate the training and confirm the skills actually acquired, an average score of 10 out of 20 or higher was required.
Learners’ levels of satisfaction with the training received
Overall satisfaction with the training received, by cohort, was as follows:
- Cohort 1: 61.9%
- Cohort 2: 84.2%
- Cohort 3: 93%
Collaborateurs clés
