Disentangling the language effect in South African schools: Measuring the impact of ‘language of assessment’ in Grade 3 literacy and numeracy

Author: Nicholas Spaull

A great majority of South African students are taught in a language that, for them, is a second or third language. Does this explain the high levels of underperformance in reading and mathematics?

This study is based on two tests that were given to 3 402 Grade 3 students from all 11 language groups. The tests were one month apart, using the same questions. One test was in the language of learning and teaching (LOLT) of the school; and the other test was in English. Grade 3 is a useful time to look at the effect of language on mathematics proficiency because students are taught in their mother tongue until Grade 4. First, the students were assessed in the language they knew best, and in which they had been taught for three years, and by teachers who were mostly teaching in their mother tongue. Then they were assessed in English.

Language is not the only factor affecting performance in reading and mathematics, however. The effect of home background and school quality – that is, parental education, teacher quality, resources, geographic location, school functionality and socio-economic status – was also considered.

Literacy test items were: Matching word to picture, Fill in missing word (cloze), Retrieve, Infer, Interpret, Evaluate, Write a sentence.

Numeracy test items were split into one of three categories based on the language content of the item, from: ‘no language content’ to ‘ambiguous’ (where the question had some language content but could be solved by deductive reasoning without any understanding of the language) to ‘high language content item’ (if it was not possible to solve the problem without understanding the language content).

The author found that the effect of working in your own language is equivalent to one or two years’ learning in literacy or numeracy. But the effect of disadvantaged home background and poor school quality is nearer to four years.

The study concluded that the ‘language effect’ is not the most important factor in determining achievement in South Africa. The authors do not negate the importance of language, but rather stress the importance of the quality of instruction.

Full Text


1 thought on “Disentangling the language effect in South African schools: Measuring the impact of ‘language of assessment’ in Grade 3 literacy and numeracy”

Leave a Comment

Your email address will not be published. Required fields are marked *