Longitudinal predictors of word reading development.

Abstract

The authors review the scientific evidence regarding the prediction of reading development from six cognitive constructs: phonological awareness, phonological decoding, naming speed, orthographic processing, morphological awareness, and vocabulary. The authors describe the theoretical connexions amongst these indicators and word reading. The authors conclude that there is substantial evidence of each contributing to reading, and in most cases, this contribution is above and beyond that of the other predictor constructs. These findings have implications for early and regular cognitive assessment, and for curriculum development and teacher education. Keywords: reading development, phonological awareness, naming speed, orthographic processing, morphological awareness Our goal in this paper is to review evidence concerning six major factors that have been shown to be related to reading development. We chose to focus on cognitive factors, not because motivation and self-perception are unimportant, but rather because the cognitive factors are more plausible causes of reading development. We also confined our discussion to specific cognitive factors that are involved in the process of reading, rather than on more general cognitive factors, such as intelligence, that are related to a broader range of behaviours, and whose specific causal role is more difficult to establish. Similarly, we omitted the various demographic characteristics that have been related to reading, such as gender and socioeconomic status (SES), partly because it is difficult to establish their causal status or exercise an influence on them, but also because, like the socioemotional factors, they operate on reading through the cognitive mechanisms that support reading. Finally, we focus on word reading, and not on reading comprehension. We do this partly because of the evidence that reading comprehension depends upon word reading (e.g., Adams, 1990), and partly because the cognitive factors on which we focus have a more proximal relationship to the more basic aspects of reading. We seek to understand the longitudinal predictors of reading for two reasons. First, because reading is a critical academic skill and because it takes so long to acquire, we would like to be able to predict reading difficulties before they occur and act to prevent them as best we can. Second, because the cognitive processes involved in learning to read are diverse and causally tangled, we explore the development of reading and its underlying processes so that we can untangle these causal relations. The first justification should lead to better educational practise, the second to better theory and practise. Reading instruction, broadly speaking, includes the intended curriculum, classroom teaching methods, assessment practises, remedial practises, and so on. These practises tend to develop in an ad hoc manner, with relatively little input from theory. Curriculum planners and classroom teachers are often not in a good position to judge theories and the evidence upon which they are based. Therefore, we see it as the responsibility of researchers in psychology and educational psychology to provide the empirical research foundation on which best practises in curriculum, teaching, assessment, and diagnosis are designed. A comprehensive theory based on solid evidence would provide a better basis for the development of principled practise and could make a substantial contribution to teacher education. The remainder of this paper consists of three sections. In the first, we briefly review several theories and sketch an overall theoretical framework that involves six causal factors. In the second, we examine each of the six factors in greater detail, and the evidence for their effects on reading. In this section, we pay particular attention to how the causal factors are related to each other, and whether they have unique effects in the presence of each other. …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 177 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.