FEEDBACK: 'Do children understand what they're reading?' Yes, when they've been taught to read, say the authors of a recent Houston report on the subject.
Originally printed in the Toronto Globe and Mail on March 18, 1997.
'Do Children Understand What They're Reading?' This headline on the March 3 commentary by Professors David Booth, Curt Dudley-Marling, Sharon Murphy and Gordon Wells, on the subject of our recent study of reading comprehension, is the right questions to ask
if the reading wars are ever to be resolved. The answer is indeed complex and cannot be answered by only one study, as Prof. Booth et al. point out.
Yet our study is part of a body of research described in a Jan. 25 [1997] editorial in The New York Times that shows the importance of accurate and fluent decoding to reading comprehension. In short, there are few good comprehenders of text who are not also good decoders of words.
Many children from literate households will induce the alphabetic code on their own and will thrive in the literature-first approach of whole language. However, for children at risk for reading failure because of socioeconomic disadvantage and low achievement (i.e. children eligible for Title 1 services in the United States), as well as children who have biological risk for reading problems, the connections between letters and sounds warrant explicit instruction.
In our research with first and second grades receiving Title 1 services, we found that the more explicit the instruction in the alphabetic code, the more accelerated the rate of improvement in decoding and overall reading achievement over a two-year period. This finding was particularly apparent among students with the lowest reading ability.
We are not "pure phonics advocates," but advocates of balanced, complete approaches to reading instruction. We wonder whether Prof. Booth et al. have read our study. They were also incorret in saying there were no differences in reading comprehension in our studies. We gave two tests of reading comprehension. The groups receiving direct instruction in the alphabetic code had significantly greater reading comprehension thatn the literature-emphasis groups. These results are not surpriseing, given the need for decoding to be sufficiently automatic that memory and attention can be devoted to grasping the gist of the text.
Concerning the professors' criticism that we are not competent to judge what is and is not a whole-language program, we would point out two facts. First, the school district in which our study was based is considered by whole-language proponents to be a model of whole-language implementation. Second, our study consisted of two groups of children receiving whole-language instruction. One group received instruction from teachers trained solely by the school district. The other group received auxiliary training by doctoral-level staff from our project hired because of the expertise in whole-language instruction
Children whose teacher had been trained by our staff scored higher on average than the other group of children receiving whole-language instruction, though not significantly so. Moreover, both groups performed more poorly on measures of decoding and comprehension than children receiving the highest level of direct instruction in the alphabetic code.
We share the concern of Prof. Booth et al. that comprehension be the goal of reading instruction. We appear to differ, however, on teh value attached to explicit instruction in the phonological, alphabetic and orthographic (spelling) skills necessary to comprehend print.
Our differences are not simply characterized as the bottom-up skills approach of phonics versus the top-down meaning approach of whole language portrayed in the media. Our differences relate to the instructional practice around how children learn sound-symbol relations.
Prof. Booth et al. say that "children learn about these relationships as they engage in reading and writing real, continuous texts." We concur with this statement with respect to becoming fluent decoders. However, there is a period during beginning reading instruction when all children benefit from practicing letter-sound connections in decodable text. To immerse children in a print-rich environment without instruction in letter-sound correspondences and practice in decodable text is to doom a large percentage of children to reading failure. The four-year study by Prof. Connie Juel (1988) showed that of children who were good readers in Grade 1, 87 percent remained good readers in Grade 4 and 13 percent became poor readers. Of the children who were poor readers in Grade 1, 88 percent remained poor readers in Grade 4 and 12 percent became good readers.
First-grade reading instruction matters. Let's make sure all first-graders become good decoders so that they will be able to
understand what they're reading.
Barbara Foorman is a professor of educational psychology at the University of Houston in Texas. David Francis is an associate professor of psychology at the University of Houston. Jack Fletcher is a professor of pediatrics at the University of Texas-Houston Medical Center.
