EdData II: Data for Education Research and Programming (derp) in Africa

South Africa grade 6 achievement by home language and province, 2005 
A five-year study (2008–2012) of a pilot program in Cameroon also demonstrated that children 
who were taught in a familiar language, Kom, performed significantly better—124% on 
average—in multiple subjects (including math and English) than a control group of peers who 
attended schools where English was the MOI (Walter & Chuo, 2012). 
Figure A2
shows the 
significant difference in learning outcomes for children in the Kom language program (KEP) and 
children in the English-medium program (English).

---

61 
Annex A 
Figure A2. 
Cameroon Kom program – 2011 results 
In Mali, students in bilingual schools (called Pédagogie Convergente schools) learn exclusively 
in their L1 in the first few years of primary education, then learn in French for half of the school 
day in grades 5 and 6. Evaluations have shown that these children consistently outperform their 
peers in French-only schools on end-of-primary school national exams (UNESCO, 2008). 
In Vietnam, educators recently developed a new curriculum to provide instruction in familiar 
languages to minority-language speakers in the country. The result was that 68% of grade 1 
learners achieved the level of “excellent” compared to only 28% of children who were not 
learning in their L1 (UNICEF, 2011). Importantly, the improvements in learning outcomes were 
true in math as well; children who received math instruction in a familiar language scored, on 
average, 75% on the assessment, while children who received instruction in an unfamiliar 
language scored only 61% (UNICEF, 2011). Similarly, an analysis of the achievement of year 8 
learners in Ethiopia between 2000 and 2004 showed that performance in mathematics and the 
science was far better for those using L1 as the MOI than for those using English. Research from 
southern Africa, too, has shown the influence of language on learning. In Botswana, pupils 
taught in Setswana had significantly better understanding of science concepts than pupils taught 
in English (Alidou & Brock-Utne, 2011).
Importantly, instruction in L1 languages that includes 
assessment 
in a familiar language allows 
pupils to better show what they have learned, leading to more accurate learning assessments 
because language is no longer a “confounding factor” in interpreting the results.
9
 This in turn 
helps teachers to better identify what children know and do not know, and consequently provide 
appropriate instruction. Moreover, use of L1 for assessment appears to be particularly beneficial 
for girls because any negative preconceptions on the part of teachers regarding girls’ academic 
ability are challenged (Benson, 2005; Hovens, 2002; Ouane & Glanz, 2010). 
9
If assessments are conducted in a language a child does not understand, poor outcomes are difficult to interpret 
because one does not know whether the child did not understand the academic content, or if s/he simply did not 
understand the language of the assessment.
0
10
20
30
40
50
60
70
Class 1 (2010)
Class 2
Class 3
Class 4