Phonemic-similarity effects in good vs. poor readers

JAMES W. HALL 0 1 KIM P. WILSON 0 1 MICHAEL S. HUMPHREYS 0 1 MARGARET B. TINZMANN 0 1 PAUL M. BOWYER 0 1 0 This research was supported in part by a research contract from the Bureau of Education for the Handicapped, Office of Education (US HEW OF 300 770 493) for the University of Illinois's Chicago Institute for Learning Disabilities. We are grate ful to the school officials , teachers, and children in Evanston and Mt. Prospect whose cooperation made this research possible, and to the Wisconsin Center for Education Research, University of Wisconsin-Madison, which is supported in part by a grant from the National Institute of Education (Grant NIE-G-0009), for assistance in preparatio n of this manuscript. Michael S. Humphreys now is at the University of Queensland, and Kim P. Wilson is at the University of Michigan. Requests for reprints should be addressed to James W. Hall, Department of Psy chology, Northwestern University , Evanston, Illinois 60201 1 Northwestern University , Evanston, Illinois Experiments 14 examined immediate serial recall of rhyming and nonrhyming items by normal and poor readers in Grades 2-4. Children with generally low achievement were excluded from the poor-reader groups, so that the achievement deficit of the poor readers was centered in reading. The poor readers did not differ from the normal readers in their susceptibility to phonemic similarity either with letter lists or with word lists. Children low in both achievement and intelligence were included in Experiment 3, and they also showed normal susceptibility to phonemic similarity, except that phonemic-confusion effects were reduced when task-difficulty levels were high. Experiment 5 further demonstrated that the serial-recall task is relatively insensitive to phonemic-similarity effects when difficulty levels are high. Previous results suggesting that poor readers are relatively insensitive to phonemic similarity in such tasks may have been an artifactual consequence of marked differences in overall task difficulty for the groups compared. Implications of variations in sample-selection procedures also are discussed. - particular population of poor readers those results may be generalized. In the Shankweiler et al. experiments and in many, if not most, other studies concerning children with reading deficiencies (or reading disability or dys lexia), interest seems to be in a particular subpopula tion of poor readers. That subpopulation consists of individuals with an achievement deficit specific to, or at least primarily restricted to, reading. This specific focus of interest is not always made explicit, but generally may be inferred by the investigator's attempt to exclude from the sample of poor readers those individuals with generally low intellectual ability. The idea seems to be to isolate and study those individuals (generally children) whose performance in decoding written words is low relative to their performance on other sorts of intellec tual tasks. Given that the focus of interest in this research is on reading ability, it seems surprising that in very few such studies is achievement in other skill areas taken into account in sampling the poor readers selected for study. That is, in nearly everyone of the many dozens of studies we have examined, the only achievement data systematically taken into account are reading-achieve ment data. Given the substantial relationship between reading achievement and achievement in other intel lectual skills, one can be reasonably certain that samples of poor readers selected in this way will include children whose achievement difficulties extend to areas other than reading. In fact, with such procedures, a sample of poor readers could well include some children whose achievement is lower in mathematics, for example, than in reading. In short, such sampling procedures seem somewhat inconsistent with the goal of understanding the source of deficits centering in reading (word decoding) and leave ambiguity regarding the particular sub group as for the other. If this is correct, then it calls population of poor readers to which any result may be into question the results of a second series of experi generalized. ments, this time with word lists and sentences, in which For the above reason, we designed our initial experi susceptibility to phonemic similarity appeared to be ments to look specifically at children whose achievement lower for poor readers than for good readers (Mann deficits were centered in reading. This was done by et al., 1980). In the Mann et al. experiments, there was no considering achievement in mathematics, as well as obvious floor effect of the sort found in the Shankweiler performance on a test of general intelligence, in the et al. experiments, but a very great superiority of good selection of low reading samples. In addition, Experi to poor readers on nonrhyming items was present, ment 3 included a group of poor readers whose general suggesting the possibility that the critical interactions in intelligence and mathematics achievement also were low those experiments also arose from a scaling problem. relative to their age-grade peers. This possibility will be considered further in the general Our approach to sample selection permitted us to discussion of our results. determine whether unusually low susceptibility to An additional impetus for the present experiments phonemic confusions is peculiar to children with a was that Hall, Ewing, Tinzmann, and Wilson (1981) specific reading deficit or whether it also (or instead) found equivalent susceptibility to phonemic-confusion characterizes children with more pervasive achievement effects when a small group of adolescents and adults deficits. Another possibility that our experiments who were severely retarded in reading were compared addressed is that the differences in phonemic confu with normal readers. The result was obtained under sions between the poor and good readers reported by circumstances in which scaling problems of the sort Shankweiler et al. (1979), and also by Mann, Liberman, discussed above seemed to be absent, and at the very and Shankweiler (1980), arose from a measurement least it raises questions regarding the generality of the (scaling) problem and really did not reflect differential conclusions reached by Shankweiler et aI. (1979). sensitivity of the groups to phonemic similarity. This Given the points raised above, together with the in possibility is discussed next. fluential role of the Mann et al. (1980) and Shankweiler The critical evidence for differential sensitivity to et al. (1979) studies in the development of theory rephonemic similarity in the Shankweiler et al. (1979) o garding reading deficiency, it seemed sensible to examine experiments was an interaction between groups and list this matter more fully. This is what we attempted to do type (rhyming vs, nonrhyming letters): The rhyming in the five experiments now to be reported. Experi (...truncated)