47
2018) suggests that OIC countries may be also undergoing a similar learning crisis. Even after
several years in school, the vast majority of students lack basic literacy and numeracy skills. In
recent assessments in urban Pakistan, only three-fifths of grade 3 students could correctly
perform a two-digit subtraction; in rural Pakistan, just over two-fifths could (WDR 2018). Across
51 countries which includes OIC member states such as Nigeria, Pakistan and Bangladesh, only
about half of women who completed grade 6 (but no higher) could read a single sentence. As a
matter of fact, it is predicted that 40 percent women would be illiterate even if all women
completed at least six years of primary schooling (Pritchett and Sandefur 2016). Further
evidence on the weak relationship between schooling and learning has emerged based on the
Financial Inclusions Insights (FIIs) survey data on 10 countries including OIC members such as
Bangladesh, Pakistan, Indonesia, Nigeria, and Uganda (Kaffenberger and Pritchett 2016).
Countries on each high quality nationally representative data on student performance is
unavailable also face the basic challenge of bringing children to school. According to the WIDE
database, less than fifty percent children complete four years of education in OIC countries such
as Afghanistan and Senegal. More than half of the world’s out-of-school children live in just 15
countries. Yet, these include 7 OIC member countries namely, Nigeria, Pakistan, Bangladesh,
Niger, Yemen, Burkino Faso and Mozambique (UNESCO 2014).
2.1.6. The Determinants of Student Achievement
In this section, we build on the earlier descriptive analysis of the variation in student
achievement in PISA and other international assessments and study the correlates of learning
outcomes using multivariate regression analysis. Table 2.1 presents the OLS estimates of the
determinants of student achievement PISA 2012 in the OIC, OECD and non-OECD countries. Data
on all 9 OIC countries are pooled as a single population group. For comparison purposes, the
results are also reported for participating OECD and non-OECD countries. Individual student
level score in mathematics, reading and science are used separately as dependent variables so
that for each 3 groups of countries, we present three regression models. The factors influencing
student performance in OIC countries clearly differ when compared to OECD and non-OECD
(and non-OIC) countries.
Key findings are as follows:
 The wealth effect is relatively higher in OIC countries. The coefficients onwealth quintile
dummies are much larger compared to those for OECD sample.
 The gender gap in reading (to the disadvantage of boys) is also much larger in OIC
countries.
 Teacher shortage does not display any systematic influence in OIC countries; only in
case of math the coefficient is significant while this variable also displays a negative and
significant influence in OECD sample.
At the same time, there are a number of common drivers of student performance when
compared to OECD countries. Among child-specific factors displaying a positive influence:
 Pre-school attendance matters regardless of test subjects in OIC as well as OECD
countries.
 There is also a significant and positive correlation between private school attendance
and student achievement. This is particularly pronounced in case of OIC countries.
Among school-specific factors displaying a positive influence are:
 Average disciplinary climate in school
 Number of computers available
 Proportion of certified teachers
 Autonomy over school budget