Exploring the Nature and Transfer of Orthographic Learning within the Self-Teaching Hypothesis

Abstract

The transition to being a skilled, fluent reader is an important focus of early education and, as such, of research on reading development. A prominent theory, the self-teaching hypothesis (Share, 2008), describes how this transition occurs. Share proposes that children learn the spelling patterns of new words during independent reading, helping to build a large store of known words that is essential for efficient word recognition and, in turn, reading comprehension. In this dissertation, we test several questions regarding the nature of self-teaching and how it supports reading development. Specifically, three studies test four core predictions of the self-teaching hypothesis: beginning readers are capable of self-teaching, accurate phonological decoding is required for self-teaching, self-teaching occurs on a word-specific basis, and self-teaching results in long-lasting high quality orthographic representations. Results from Study 1 show that beginning readers in Grades 1 and 2 engage in self-teaching, although it is less durable for those in Grade 1. Other findings from Study 1 suggest that self-teaching in beginning readers does not require accurate phonological decoding nor is it strictly word-specific, with children transferring learning of one word (e.g., feap) to facilitate processing of related novel words (e.g., feaper). Findings from Study 2 suggest that self-teaching is not a strictly word-specific process for more experienced readers in Grades 3 to 5 either. Here I show that learning one word (e.g., feap) facilitates better learning of related words (e.g., refeap) and, in turn, these accumulated learning experiences lead to even better processing of additional novel words (e.g., misfeap). Lastly, findings from Study 3 suggest that self-teaching leads to long-lasting quality orthographic representations in more experienced readers, with Grade 5 children able to access both orthographic and semantic information one year after learning. I place these findings in context of the original predictions of the self-teaching hypothesis (Share, 2008), and conclude by proposing an updated theoretical framework. This new framework provides three principles that may more accurately describe self-teaching and its role in reading development for beginner readers and more experienced readers alike.