Word and Person Effects on Decoding Accuracy

Abstract

The purpose of the current study was to extend the literature on decoding by bringing together two lines of research, namely person and word factors that affect decoding, using a crossed random-effects model. The sample was comprised of grade 1 students who were at risk for developing reading difficulties. A researcher-developed pseudoword list was used as the primary outcome measure. Because grapheme-phoneme correspondence knowledge was treated as person and word specific, we are able to conclude that it is neither necessary nor sufficient for a person to know all grapheme-phoneme correspondences in a word before accurately decoding the word. As predicted, students with lower phonemic awareness and slower rapid naming skill have lower predicted probabilities of correct decoding than their counterparts with superior skills. Results also reveal that words are more difficult if they contain an infrequent as compared to a frequent rime and a complex vowel as compared to a complex consonant. By assessing a person-by-word interaction, we found that students with lower phonemic awareness skills have more difficulty applying knowledge of complex vowel graphemes compared to complex consonant graphemes when decoding unfamiliar words. Implications of this methodology and of the results are discussed in light of early reading instruction and future research.