Rational analysis of reading: Identifying optimal representations and the implications for reading instruction

One of the difficulties in learning to read English is that the relationship between spelling and sound is not always consistent (eg 'ck' is always pronounced as in tick, but 'ou' can be pronounced several ways; touch, group, mould, etc.). Larger spelling-to-sound mappings can often reduce this inconsistency (eg 'oup' is always pronounced as in group). A word can therefore be decoded in different ways, depending on the size of spelling-to-sound mappings used. Whilst research supports the use of spelling-to-sound mappings as vital for learning to read, opinion has been divided over what type, and how many taught mappings produce the best results in terms of reading ability. The primary goal of this research is to identify which spelling-to-sound mappings have the greatest potential in terms of learning outcome for beginning readers. It uses methods developed within psychology and Information Theory to evaluate issues such as whether some types of spelling-to-sound mappings (eg small as in 'ou' or large as in 'oup') are in general more useful than others, which mappings would appear in an optimal set for instruction purposes, and which mappings should be introduced early or late in instruction.