Thompson-McKay-2008

Abstract

Two studies were conducted across three countries to examine samples of beginning readers without systematic explicit phonics who had reached the same level of word reading accuracy as comparison samples with high and moderate explicit phonics. Had they employed any compensatory learning to reach that level? Four hypotheses of compensatory learning or performance were tested on the samples, all of which represented the lower half of the normative distribution of word reading accuracy. The two samples without explicit phonics received teaching that centered on story text reading and some receptive phonics that arose from this text reading. They did not compensate by relatively greater use of a larger psycholinguistic grain size in the form of rime units. Nor did they compensate by trading off comprehension for text word reading accuracy. In a microtraining study, they showed no compensation in proficiency of initial learning of lexical orthographic representations. For all samples, this initial learning was less effective with spelling than reading training trials. In reading text, the samples without explicit phonics did not compensate by trading off speed for accuracy, or comprehension. On the contrary, they read text faster than the explicit phonics samples. The extra classroom instruction time available to them for text reading, with the consequential extra exposure and practice of word reading, would explain this result.

Excerpts from Full Text:

Also in Study 1 it was found that children without explicit phonics did not trade off text reading speed to achieve equal word reading accuracy and equal reading comprehension/recall. On the contrary, across the samples, children’s oral reading of text was faster without explicit phonics
instruction. Oral reading speed, adjusted for total word reading accuracy, was 33.5 words per minute for the mean of the two nil samples. On average they read 46%more words per unit of time than the high phonics and 20% more than the moderate phonics sample (calculated on the values in Table 2). This was for a word accuracy level for the texts that averaged 80%.

 

We have eliminated these alternatives and are left with a result showing that,despite equal word reading and reading comprehension accuracy, the speed of reading text was faster with less explicit phonics instruction, in readers representing the half of the population below the normative mean.

As we observed in the introduction, among the studies with an adequate sample size reviewed by the National Reading Panel (2000, chap. 2, pt. 2) on first-grade classrooms or small groups, none was reported to have a measure of text word reading accuracy or speed.

 

There is a second explanation, which does not conflict with the evidence
reviewed by the National Reading Panel (2000). Given that the same total reading instruction time is available, a teaching program that includes explicit phonics instruction separate from story text reading cannot provide as much text reading instruction time as one in which text reading dominates and teaching of phonics is of the receptive type that arises from text reading. Moreover, according to the results of Martin-Chang & Levy (2005), young children’s speed of oral reading of text increases as a result of practice with the syntactic and semantic relations among the words of text, as well as with practice in reading the component isolated words. Furthermore, there is a multiplier effect of practice time on reading speed.

 

However, we have shown that with word reading accuracy and reading comprehension equal, speed of reading text was faster among children without explicit phonics. These children received some receptive phonics that arose from their text reading. Our results for children without explicit phonics were replicated over two countries, and confirmed two similar findings from the only other studies in the literature. This finding on speed of reading text is of potential significance for children’s acquisition of reading fluency,