This study examined relations between self-efficacy and outcome expectancy beliefs and spelling and writing performance. Perceptions about spelling and writing were assessed in 258 collegeage participants. Spelling performance was measured through a 50-item spelling test and writing performance by a holistically scored writing sample. The most highly correlated variables included spelling outcome expectancy and writing outcome expectancy, spelling selfefficacy and writing self-efficacy, spelling performance and spelling self-efficacy, and spelling and writing performance. A causal model relating perceptions, spelling performance, and writing performance was proposed and its appropriateness estimated. Direct effects on spelling were found for spelling self-efficacy, while spelling self-efficacy had indirect effects on writing performance and spelling had a direct effect on writing performance. The causal model was discussed in terms of changing conceptions of writing instruction and traditional views of the role of spelling as a necessary component of good writing.While spelling never has occupied more than a fraction of the attention devoted to reading, good spelling nonetheless has been regarded by many parents and educators as an essential goal in the development of literacy. Traditionally, spelling instruction involves time set aside daily for spelling activities: these include pretesting, study, and post-testing of words on predetermined lists; a systematic method of study; drill and practice of isolated words; and activities that encourage students to write the words in a meaningful context (see review by Brown, 1990). This bottomup philosophy (Yellin, 1986), emphasizing rules, drill, repetition, and memorization of word lists, has its roots as far back as the late 1700s when Noah

This study examined relations among spelling performance and students' beliefs about spelling, including self-efficacy for spelling ability, outcome expectancy for spelling, and attributions for good spelling across grades 4,7, and 10. Spelling self-efficacy remained relatively constant across grades. Spelling outcome expectancies for adult life and school declined across grades, as did effort and ability attributions for spelling success, with a disproportional decrease in ability attributions between grades 4 and 7. Self-efficacy was the strongest predictor of spelling performance at all grade levels; attribution for ability entered into the regression for grade 4 students, while outcome expectancies for school and writing were more important in grades 7 and 10. Cluster analyses on the grade 10 sample showed that students with high efficacy as spellers and high outcome expectancy of spelling for writing were the best spellers, with the highest performance reserved for those who attributed good spelling more to effort than ability. The impact of spelling instruction on developing beliefs is discussed.Although the area of spelling has never captured the attention of educators and researchers to the degree that reading and writing have, over the past two decades there has been mounting interest in understanding the cognitive processes that underlie spelling and in integrating this knowledge with the larger body of knowledge of cognitive science (see Frith, 1980, and the review by Brown, 1990). As a maturing field, cognitive science in general has broadened its scope beyond the more 'purely cognitive' models proposed initially (Anderson, 1983;Rumelhart, 1975;Schank and Abelson, 1977;Winograd, 1975) to include a variety of affective and motivational variables affecting learning (Bandura, 1982(Bandura, , 1986 Dweck and Legget, 1988;Pintrich, We gratefully acknowledge the contributions of Greg Schraw, University of Nebraska-Lincoln, for his feedback and helpful recommendations relative to data analysis and preparation of this paper.

No abstract

No abstract

The role of decision making in biasing recall of prose was examined. Both making a decision subsequent to reading and encountering an experimenter-provided decision resulted in recall biased in the direction of the decision. Conditions in which subjects made their own decisions, however, resulted in greater levels of decision-congruent recall than did conditions in which the experimenters furnished decisions to subjects. In addition, the level of recall of incongruent information in the subject-made decisions conditions was as high as the level of congruent information observed in the experimenter-provided decisions conditions. A memory-restructuring model, suggested by Dellarosa and Bourne (1984), accounts for the results if it is true that greater elaboration of both decision-congruent and decision-incongruent information during subject-made decisions occurs.Recently, Dellarosa and Bourne (1984) observed that the provision of additional information subsequent to reading had essentially the same effect on bias in recall as did decision making. They argued that the decision-making process itself had little impact on how information about the passage was stored. They suggested that the decision served the same function as a statement given readers subsequent to reading; that is, either the decision itself or the statement was stored as afocal point for the information retained in memory related to the passage.Although it seems clear from Dellarosa and Bourne's (1984) work that both decisions and statements given after reading bias recall, it is unclear whether the processes involved in obtaining these biases are the same. Several studies have shown that decision making increases the amount of text material recalled relative to other procedures that do not involve decision making on the part of readers (e.g., Benton, Glover, & Bruning, 1983; Benton, Glover, Monkowski, & Shaughnessy, 1983;Glover, Bruning, & Plake, 1982). It may very well be that statements provided after reading bias recall by reducing the potential that decision-incongruent information will be available at the time of recall. In terms of decisions, however, it seems that an alternative explanation is possible; that is, biases may result not from the attenuated availability of decision-incongruent information at the time of recall, but rather through an increase in the amount of decision-congruent information available.The current study examined the effects of decision making and the provision of statements subsequent to reading on biases in readers' recall. Five conditions were included: subject-made positive decisions (SP), subjectmade negative decisions (SN), experimenter-provided positive intervening information (EP), experimenterprovided negative intervening information (EN), and conRequests for reprints should be sent to John A. Glover, Department of Educational Psychology, University of Nebraska, Lincoln, NE 68588-0440. Copies of the stimulus materials are available on request. 77 trol. Both positive and negative options were included becaus...