Self-report and test discrepancy evidence from national literacy survey in Bangladesh

Abstract

This paper aims to explore the discrepancy between self-reported and test based literacy estimates. The Education Watch national literacy survey data of 2002 were used to do so, where literacy status of a nationally representative sample of 13, 145 persons was collected through the above two different methods. The findings revealed that the literacy rate generated through a literacy test was significantly lower than that found through self-report method. At the national level the amount of discrepancy was 9.5 percentage points. The level of discrepancy varied from one group of population to another. Discrepancy was more likely among the primary school educated rural females of age 15-24 years and less likely to never schooled and college educated urban males of age 25 years and above. This shows that literacy assessment through self-report method cannot be equally appropriate for all population. More risk especially due to over reporting occurred among those having some years of schooling. A question of quality of primary education also raised here. Considering the low quality of school education and increased enrollment in Bangladesh this paper suggests for a paper-pencil based literacy assessment rather than oral reporting.