Jeffrey J. Walczyk

How do the efficiencies of reading subcomponents relate to looking back in text

The compensatory-encoding model of reading suggests that experienced readers compensate metacognitively for inefficient reading subcomponents or cognitive resource limitations. For instance, readers with less efficient access to information in working memory are predicted to look back in text more than those with more efficient access to information in working memory. Both groups of readers still have good comprehension. One hundred nine adult readers were measured on the efficiencies of selected reading subcomponents. Texts were read from a computer screen 1 sentence at a time. The extent to which readers looked back in text (reread sentences) was noted. Those with more efficient access to information in verbal working memory looked back less, consistent with the model. Theoretical and educational implications are discussed.

The relation between low- and high-level reading skills in children

Abstract Little research has explored the relation between low- and high-level reading skills in children. Efficient low-level skills, such as lexical access, may afford savings of limited attention and working memory so that attention-demanding reading activities proceed unencumbered ( Perfetti, 1985 ). Second, fourth, and sixth graders were tested on measures of low-level reading efficiency: the efficiency of decoding, the efficiency of semantic memory access, and verbal working-memory span. Children were also tested on measures of high-level comprehension outcomes: error detection, a test of inferencing, and the number of spontaneous inferences made while recalling a story. The results indicate that high-level comprehension does not depend on the efficiency of lexical access processes across grade levels. Verbal working memory span, however, is important in high-level comprehension. Those with larger verbal memory spans may have ready access to information for high-level comprehension.

Are General Resource Notions Still Viable in Reading Research

To explore the interplay between reading subcomponent efficiency and comprehension processes in terms of resource competition, 2 experiments were conducted. In Experiment 1, adults were tested on the efficiency of lexical access, semantic memory access, verbal working memory span, contextual priming, and the efficiency with which anaphors are resolved. This profile was used to predict the efficiency of high-level inference generation when readers control the pace of reading. In Experiment 2, this profile was used to predict the efficiency of high-level inferencing as text was presented at a preset rate. The data suggest that when readers control reading rate, they compensate for subcomponent inefficiency such that high-level comprehension is unrelated to this inefficiency

A developmental study of children's ability to adopt perspectives and find errors in text

Abstract Research on comprehension monitoring using the error detection paradigm indicates that children have difficulty monitoring their comprehension of text. In past studies children have not been specifically asked to apply their knowledge while listening to or reading text. Consequently, children may not have been listening or reading for meaning. Perspective taking is an example of a task which can encourage children to apply what they know to information they are encoding. In the present study, two stories were used, each of which could be read from two distinct perspectives. Each story contained information relevant to each perspective. In the first of two studies it was shown that second- and fourth-graders can successfully rate passage content from each of two perspectives. In the second study, two errors were placed in each story, one relevant to each perspective. Second- and fourth-graders were asked to adopt perspectives and then listen to each story. Error detection was recorded. In addition, subjects ware asked to recall the stories. Fourth-graders detected more perspective-relevant errors and recalled more perspective-relevant information. It is argued that, while second- and fourth-graders have the knowledge necessary to identify errors, fourth-graders are more likely to apply it on-line during comprehension.

The development of verbal efficiency, metacognitive strategies, and their interplay

This article explores the interplay between low-level reading subcomponents, metacognitive aspects of reading, and how this relation changes developmentally. First, theories relevant to the question of how subcomponent efficiencies, available resources, and metacognitive aspects of reading interrelate are reviewed. Second, research documenting increases in a readers subcomponent efficiencies is briefly surveyed. Next, research revealing the childs developing metacognitive competencies is explored. To conclude, compensatory-encoding theory (Walczyk, 1993) is used to propose a model of the development of the interplay between resources, subcomponent efficiencies, and metacognitive aspects of reading. Diagnotic and intervention implications are considered.

Does activation of a script facilitate lexical access

Abstract According to script-based accounts of text processing (e.g., Walker & Yekovich, 1987 ), activation of a script during reading can facilitate lower level processes, such as lexical access, in a top-down fashion. A competing perspective, known as modularity theory Fodor 1983 suggests that low-level processes, such as lexical access, can become informationally encapsulated and thus are not influenced by script activation. In two experiments, passages were read off of a computer screen. For each of four passages six words were identified as either central or peripheral to the script. Versions of each passage were constructed in which these same six words were not related to the script. Thus, these words appeared in both scripted and nonscripted versions of each story. The efficiency of lexical access was measured in two ways: by either word reading time or naming latency. The data, collected within subjects, provide little evidence of script-related facilitation of lexical access, thus supporting modularity theory.

A computer program for constructing language comprehension tests

Abstract The sentence verification technique (SVT) has become increasingly popular among educators and psychological researchers as a powerful means for constructing cost-effective, psychometrically valid tests of literal language comprehension. Instead of testing a reader or listener on a few details of information in a passage, an SVT test assesses comprehension over an entire text. This article reviews the SVT technique, why it was devised, and some of its advantages over traditional measures, and then discusses a program, now available, that automates and greatly simplifies the process of constructing SVT tests. The program is very powerful, and includes such features as an internal text editor and a measure of text readability.

A program for constructing SVT tests: An alternative way of assessing text comprehension

The sentence verification technique (SVT), developed by Royer, Hastings, and Hook (1979), provides developmental and educational researchers with an alternative way of assessing text comprehension based on all of the information contained in a passage. Students must decide if test sentences have the same meaning as sentences that appeared in an original passage. The program described here automates the more laborious aspects of constructing SVT tests. The user must create a passage using a word processor. The program will prompt the user for information as needed. Finally, printable text files (SVT test and answer key) are generated. The program is written in BASIC for IBM and IBM-compatible computers.