Language Development and Learning to Read – Diane McGuinness Page 3

In my research on children's reading strategies (McGuinness 1997a), I found that by the end of first grade, children in whole-language classrooms were using three different decoding strategies. A small minority were decoding primarily by phonemes (one sound at a time). Another group, whom I call "part-word decoders," searched for recognizable little words or word fragments inside bigger words. A third group ("whole-word guessers") decoded the first letter phonetically, then guessed the word by its length and shape – the overall visual pattern made by the letter string. Very few children used a pure sight-word strategy (the telephone-number strategy*), and the children who did usually stopped reading before the end of the school year.

Reading test scores reflected these strategies, with the phonemic decoders superior, part-word decoders next, and whole-word guessers the worst. When these children were followed to third grade, the whole-word guessers had not changed their approach and were the undisputed worst readers in the class. Some part-word decoders had graduated to phonemic decoding, but the majority of the third graders remained primarily part-word decoders. Once more, phonemic decoders were far and away the best readers. This shows that children are active learners, and when confronted with vague or misleading guidelines for how to read, they try out strategies to overcome this difficulty. The fact that these strategies are different, and that they tend to stay constant over such a long period of time, is strong evidence against a developmental explanation.

* Whole word methods lead some children to believe that they can memorize each word as a random string of letters. This makes learning to read exactly like trying to memorise the telephone directory.