Marit S. Samuelstuen

Examining the validity of self-reports on scales measuring students' strategic processing.

BACKGROUND Self-report inventories trying to measure strategic processing at a global level have been much used in both basic and applied research. However, the validity of global strategy scores is open to question because such inventories assess strategy perceptions outside the context of specific task performance. AIMS The primary aim was to examine the criterion-related and construct validity of the global strategy data obtained with the Cross-Curricular Competencies (CCC) scale. Additionally, we wanted to compare the validity of these data with the validity of data obtained with a task-specific self-report inventory focusing on the same types of strategies. SAMPLE The sample included 269 10th-grade students from 12 different junior high schools. METHODS Global strategy use as assessed with the CCC was compared with task-specific strategy use reported in three different reading situations. Moreover, relationships between scores on the CCC and scores on measures of text comprehension were examined and compared with relationships between scores on the task-specific strategy measure and the same comprehension measures. RESULTS The comparison between the CCC strategy scores and the task-specific strategy scores suggested only modest criterion-related validity for the data obtained with the global strategy inventory. The CCC strategy scores were also not related to the text comprehension measures, indicating poor construct validity. In contrast, the task-specific strategy scores were positively related to the comprehension measures, indicating good construct validity. CONCLUSION Attempts to measure strategic processing at a global level seem to have limited validity and utility.

Do students' self-efficacy beliefs moderate the effects of performance goals on self-regulatory strategy use?

We examined whether perceived self-efficacy moderated the relationship between performance goals and self-regulatory strategy use in two different samples of 178 and 108 Norwegian post-secondary students. Using multiple regression with interaction terms, we found that perceived self-efficacy moderated the relation between performance-avoidance goals and reported use of self-regulatory strategies for students in a competitive, performance-oriented context. Specifically, in that context, there seemed to be a negative effect of increased performance-avoidance goal orientation for students with high self-efficacy and a positive effect of increased performance-avoidance goal orientation for students with low self-efficacy. While the nature of this moderator effect is not consistent with what has previously been suggested by researchers using a goal orientation framework, our findings point to the importance of examining self-efficacy moderator effects in different study contexts.

Psychometric properties and item-keying direction effects for the learning and study strategies inventory-high school version with Norwegian students

The purpose of this study was to examine whether the Learning and Study Strategies Inventory-High School Version (LASSI-HS) would yield reliable and valid scores when using a Norwegian translation with a sample of Norwegian high school students. In addition, the author examined possible response biases associated with item-keying directions within the LASSI-HS. Three alternative factor models proposed in prior research were evaluated, with results yielding support for the model suggested by Olaussen and Bråten. However, method effects associated with item wording were found, and positively worded items were characterized by more response biases than negatively worded items for the Norwegian students.

Eye-movement parameters and reading speed

This study addressed the relationship between four eyemovement parameters and reading speed of 20twelve-year-old children during silent and oralreading. The results indicated that each of theparameters correlated significantly with speed ofsilent reading. In fact, reading speed could bepredicted by the following variables: recognitionspan, average fixation duration, and number ofregressive saccades. Moreover, the relationshipbetween silent and oral reading speed, on one hand,and between reading fluency and reading errors in oralreading speed, on the other, was examined. The resultsindicated that in terms of reading speed, significantinterrelationships existed between silent and oralreading. Furthermore, fluency and errors wereintercorrelated phenomena, but only fluency wassubstantially related to oral reading speed.

Poor Readers-Good Learners: A Study of Dyslexic Readers Learning with and without Text.

Our purpose was to examine how high-achieving dyslexic readers compensated for their poor decoding skills both during independent learning from text and in the broader learning context of home and school. The participants were 8 Norwegian junior high school students who had performed well in school despite diagnosed difficulties with single word decoding and phonological processing. Through a combination of quantitative reading and questionnaire data and qualitative research interviewing, it was found that most of the students seemed to use comprehension strategies effectively during text learning to compensate for their word-level difficulties. Additionally, the participants drew on a variety of personal, social, and technological resources to compensate for their poor decoding skills when learning at home and school.

Cognitive and linguistic predictors of text comprehension

The problem was concerned with the extent to which cognitive/linguistic functions and domain-specific strategies could predict readers’ text comprehension when working with science texts. Six hypotheses were stated and tested in two prediction studies (N=101 eighth-graders), and one experiment (N=8 pupils of same age). Handal’s reading test was used for classifying pupils into good and poor readers. Two texts, one with a known subject and another with an unknown theme were imposed pupils as science reading tasks. Questionnaires and tests concerning cognitive and linguistic functions were used as independent variables in the prediction studies. Two different tests were used as independent variables (for text comprehension) in all of the three studies, and one additional criterion test (writing essays on main ideas in the texts) was applied in the experient. The experiment was organized as a 2x2 factorial design where text type and reading skill were the factors. According to the results, 5 of the 6 hypotheses could not be rejected, i.e., general concept information was the most significant predictor of science text comprehension. Furthermore, text based on a subject known to the reader always surpassed text with an unknown subject as to text comprehension. Similarily, across-domain strategies were more closely related to reading comprehension than domain-specific strategies.RésuméLa question posée était de voir dans quelle mesure des fonctions linguistiques cognitives et des stratégies dans un domaine spécifique pouvaient permettre de prédire la compréhension de textes scientifiques. Six hypothèses ont été formulées et testées dans deux études prédictives (avec 101 élèves de huitième année) et une expérimentation (avec 38 élèves du même âge). Le test standardisé de lecture de Handal a été utilisé pour classer les élèves en bons et médiocres lecteurs. Deux textes ont été utilisés comme tâches de lecture en science: un texte portant sur un sujet connu et un autre portant sur un thème inconnu des élèves. Des questionnaires et des tests concernant les fonctions cognitives et linguistiques ont été utilisés en tant que variables indépendantes prédictives. Dans les trois études deux tests différents ont été utilisés comme variables indépendantes pour la compréhension de texte. L’expérimentation une épreuve de rédaction écrite des idées principales des textes a été utilisée, en plus, comme variable indépendante. L’expérimentation a été organisée selon un plan d’expérience factoriel comportant deux facteurs: le type de texte et les compétences en lecture. Les résultats montrent que 5 des 6 hypothèses sont confirmées, l’information conceptuelle générale étant le prédicteur le plus signifiant de la compréhension d’un texte scientifique. Par ailleurs, a compréhension des textes concernant un sujet connu des lecteurs a toujours dépassé toujours celle des textes concernant un sujet inconnu. De même, les stratégies trans-domaines ont été plus étroitement corrélées à la compréhension de lecture que les stratégies spécifiques au domaine.

The Relationship between Linguistic Characteristics and Reading/Writing Performances of Norwegian Children.

The relations between different word categories and childrens reading and writing performances were examined in twenty 9-year-old children. Results indicated for Norwegian, which is more regular than English but less regular than Finnish, that the length and the frequency of words and their interactions are factors substantially related both to childrens reading, writing time (writing velocity), and spelling performances, whereas the regularity factor affected childrens spelling only. Significant intercorrelations among reading and writing (accuracy and spelling) measures were found.