Acquisition is not spongelike:Using repetitions required to learn words to investigate influences on word recognition in Year 1 English children

How many times do children learning to read need to see printed words for the words to be reliably recognised? Reitsma (1983) demonstrated that Dutch children who had made average reading progress for six months could read words they had seen as few as four times significantly faster than similar unfamiliar words. This research has been quoted widely as suggesting that children learning to read English need similar level of exposures to learn unfamiliar vocabulary.
To investigate this claim, a small group of English Year 1 children were assessed on words they had encountered varying numbers of times in books used to teach them to read. In addition to investigating whether four repetitions were sufficient for a variety of words, the vocabulary was analysed to evaluate the relative level of repetitions required for children to reliably recognise words varying in decodability, word class and morphemic complexity. The overall sample of words needed to appear in books more than 15 times for reliable recognition. Words children could decode required significantly fewer repetitions than those beyond their decoding ability. No significant differences were found for repetitions needed by words varying in word class or morphemic complexity.
Decodable words, out of all the categories analysed, were those requiring the fewest repetitions, reliable recognition being attained within the band from 4 to 15 occurrences, and might therefore be considered as candidates for ‘spongelike acquisition’. Non-decodable words, however, did not attain reliable recognition until repetitions exceeded 40, confirming in an indirect manner the critical importance of decoding skills for children’s reading development.
Repetition of vocabulary, though, a neglected factor in research, appears to be equally essential, and the results of this small pilot study seem to warrant a larger-scale investigation. Above all, what this study has shown is that, for at least some children and some types of word, acquisition is not ‘spongelike’.