Interpreting error patterns in a longitudinal primary school corpus of writing

Abstract

This paper examines the error patterns in a small longitudinal corpus of Singaporean primary school children’s writing and argues that corpus data, although highly effective in uncovering patterns that might otherwise go unnoticed, need to be meaningfully interpreted for them to be useful in contributing to improvements in pedagogy. Such interpretations require that researchers take into account a variety of other factors in addition to textual variables, including the sociolinguistic context and the particular pedagogic practices of the local educational system.