Page 192 - Contributed Paper Session (CPS) - Volume 4

CPS2173 Felicien Donat Edgar Towenan Accrombessy et al.
2. Methodology
First, the primary school achievement trend was plotted and PCA depicted
performance profiles by groups of locations and sorted determinants of
pupils’ performance, and to inform variables included in the school
performance determinants regression.
The 2013 population and housing census informed primary school
attainment and completion. Primary school completion was calculated using
the proportion of the population aged 12 years and plus that completed at
least the primary 6 school grade. The ratio is calculated by age group and by
birthyear.
Principal components analysis (PCA) is an extremely powerful statistical
tool used to synthesize information, rather than performing a graphical
representation for each of the primary education indicators (from the
administrative source). PCA calculates main components concentrating most
of the information in the matrix of original indicators in order to distinguish
clearly the school performance of municipalities (communes), and to classify
them into shorts groups with common characteristics. The advantage is to
provide policy makers with evidence on levers to activate in order to improve
the performance of the education system by targeting the intervention based
on the evidence provided by each group of communes now recognized by
their particular profile that the PCA drew. The statistical units considered to
2
realize the PCA are the 77 communes of Benin. The variables are the
indicators of primary education. The following set of 16 variables includes:
- seven (7) ratios related to student access and performance in the
education system (Gross Enrollment Ratio (GER), Gross Admission
Ratio (GAR), Completion Ratio (TAC), Net Enrollment Rate (NER),
Promotion Ratio (PR), Drop-out ratio (DR), CEP admission ratio, the
percentage of graduates after the national exam to validate primary
school grade 6 completion,
- four (4) ratios (number of pupils per teacher ratio, number of pupils
per classroom ratio, number of pupils by instructional group ratio and
number of textbooks per pupil ratio), and
- proportions and averages (percentage of qualified teachers,
percentage of civil servants (state agents), percentage of pupils sitting
on tables-benches of 2 places at most, percentage of pupils sitting on
benches and the average number of blackboards per classroom.

2 Benin is at administrative level divided in 12 regions (the Departments). Each Department is
divided in sub-regions called “Communes”, 77 in total. The Communes are composed of
Arrondissements, and the important level of geographic administrative decomposition is the
Village or City unit (Quartier de ville). In the analysis, we preferred to do the PCA at the
Commune level, because it offer significant policy decision incentive particularly regarding the
advanced decentralization process in Benin.
181 | I S I W S C 2 0 1 9