Page 191 - Contributed Paper Session (CPS) - Volume 4

CPS2173 Felicien Donat Edgar Towenan Accrombessy et al.
In Benin, several efforts have been made to raise the level of education.
However, although perceptible progress in terms of access to primary school
education, as evidenced by the significant enrollment ratios, the country faces
more difficulties in terms of the quality of education (RESEN, 2014). The
situation of the teaching of writing, reading and mathematics in Benin reveals
gaps in quality: Benin is largely behind other comparable countries. At the
beginning of schooling, the national average scores in reading are 458.3 points
and 454.7 points in maths falling below the average of the ten countries
surveyed in 2014 by the PASEC, set at 500 points. At the end of schooling, the
national average scores in reading (523.4 points) and maths (496.9 points) are
close to the average of the ten countries.
Nevertheless, these scores remain lower than those of several comparable
countries such as Senegal and Cameroon. To help understand this problem
and to find out the right policies to achieve the SDGs, this paper aims to
analyze the factors that influence school performance in language and
mathematics among students in grades 2 and 5 of primary school. The main
lessons of this study are to inform policy makers by helping them to identify
the reasons for poor pupil’s performance in Benin and thus to better define
the corrective educational policies to be implemented. Starting from the fact
that a quality school is one where quality teaching is provided, the quality of
education is measured by the achievement of students quantifiable by their
school achievements based on their PASEC1 tests scores. Thus, the present
study exploits all the PASEC data for the years 2004 and 2014 applied to Benin,
on samples of 1705 pupils of 2nd grade and 1823 pupils of 5th grade of
primary school belonging to 273 schools. We grouped the variables available
into three categories: individual variables, family variables and those related
to the school context.
To consider the hierarchical structure of the data, because of survey
sampling design, instead of using OLS (Ordinary Least Square) linear models
that have limitations, more advanced multilevel models (Michaelowa (2000)
and Bressoux (2010)) are preferred to link inputs from the educational
production function and the outputs. The econometric analysis of the
explanatory factors of pupils' academic performance will answer the following
questions: What are the individual and school context characteristics that
contribute positively to the differences in student scores? Are there links
between the two levels of the analysis? Are the main explanatory factors of
school acquisitions the same over the two periods of analysis (2004 and 2014)?
What lessons does the interaction between some key variables of the two
levels of analysis on student achievement provide?

1 PASEC is the CONFEMEN Program for the Analysis of Education Systems. CONFEMEN is the
Conference of Ministers of Education of French-Speaking Countries
180 | I S I W S C 2 0 1 9