Reading and Writing: An Interdisciplinary Journal 14: 571–598, 2001.
© 2001 Kluwer Academic Publishers. Printed in the Netherlands.
Orthographic analogies and early reading: Explorations of
performance and variation in two transfer tasks
ROBERT SAVAGE & MORAG STUART
Institute of Education, University of London, UK
Abstract. Two experiments investigated the use of orthographic analogies in 6 year olds. In
Experiment 1, 26 children were shown CVC clue and target word pairs sharing either rimes
’ – ‘pork’), heads (‘fork’ – ‘ford’) or were controls (‘fork’ – ‘hurl’). A modest advantage
for rime-clued over head- clued targets was unreliable over by-subject and by- item analyses.
Improvements in target word reading were correlated with pretest scaffolding errors (e.g.
‘pork’ misread as ‘park’). In Experiment 2, 50 children were pretaught three clue words for
each target word before being shown words that shared either rimes (‘leak
’ – ‘peak’), or medial
vowel digraphs (‘lea
k’) – ‘bean’), or were controls (‘leak’ – ‘herd’). A modest advantage for
rime-clued over vowel digraph-clued targets was again unreliable over by-subject and by-item
analyses. Neither rime nor phoneme awareness measures were correlated with rime inference
use. Vowel, but not rime inference, was correlated with scaffolding errors. Rime detection was
the strongest predictor of reading ability, whereas phoneme segmentation was the strongest
predictor of the use of scaffolding errors.
Keywords: Orthographic analogy, Reading, Rime detection
The theory that children might use orthographic inferences from the
pronunciation of known words to derive pronunciations of unknown words
commands widespread support (e.g. Baron 1977; Ehri 1992; Frith 1985;
Goswami 1986; Goswami & Bryant 1990; Share 1995; Stuart & Coltheart
1988; Thompson, Cottrell & Fletcher-Flinn 1996). Inferential strategies in
reading acquisition are particularly attractive as they may provide a powerful
self-teaching mechanism for the many irregularities of written English (Share
1995). Much of the empirical and theoretical debate surrounding inference
use has focused upon the extent to which large orthographic rime body units
(the vowel and terminal consonant(s) of a syllable) or smaller grapheme
units are functional units when orthographic inferences are used in early
reading (Baron 1977; Duncan, Seymour & Hill 1997; Goswami 1993; Marsh,
Friedman, Welch & Desberg 1980; Thompson et al. 1996). Despite much
research, disagreement over the size of the unit employed by children in