Listening comprehension and decoding in relation to reading comprehension: An exploration of an additive model

The purpose of this study was to evaluate an additive model of reading that proposes that reading comprehension is composed of the combination of listening comprehension and decoding skills. This model was inspired by Gough & Tumner's multiplicative model (1986) but is somewhat simpler. Essentially the multiplicative model states that the addition of an interaction term predicts more variance in reading comprehension performance than the subcomponents added individually to the regression equation. Specifically, this research tested the following research hypotheses: (1) There will be no interaction between decoding and listening comprehension as related to reading comprehension. (2) Listening comprehension and decoding each provide unique variance over and above the other in relating to the variability in reading comprehension.;Reading comprehension was measured with the Passage Comprehension subtest of the Woodcock-Johnson Tests of Achievement-Revised (WJ-R) (Woodcock, 1989) in both data sets. In the first data set listening comprehension was measured with the alternate form of the Passage Comprehension subtest read to the participants. Listening comprehension was measured in the second data set with the Listening Comprehension measure of the Woodcock-Johnson Tests of Cognitive Ability (Woodcock, 1989). Decoding ability was measured for both data sets using the two decoding subtests of the WJ-R, that is Letter-Word Identification and Word Attack.;The hypotheses were tested using general linear models. The dependent variable was reading comprehension with decoding skills, listening comprehension, the interaction of these variables and age as the predictors. This model was significant. The first hypothesis was not supported. Rather, Gough and Tumner's (1986) multiplicative model was supported. The second hypothesis was supported, indicating that listening comprehension and decoding each predicted unique variance in reading comprehension over and above that predicted by the other.;In conclusion, the current study added support to the multiplicative model of reading comprehension. It also extended the ages to which the model can be applied. Suggestions were made as to how this model can be used in developing a prevention oriented reading instructional system in the schools.;This study examined 60 participants, who were previously administered measures of reading comprehension, listening comprehension, and decoding skills. The data set used in the current study was formed by combining two data sets. The first set of 23 third grade participants was collected by the author during her pilot study (Loshak, 1997). The second data set of 37 3rd through 12th graders was taken from the files of a private learning clinic. The participants ranged in age from 8 to 18 years and were in grades 3 through 12. Males and females were represented in similar numbers in the sample. There was a cross section of reading abilities. Participants in the first data set read above grade level on average, and participants in the second data set read at grade level on average.