Monica Rosén

Gender differences in structure, means and variances of hierarchically ordered ability dimensions

Abstract The study investigates gender differences in cognitive abilities using a multivariate, latent variables approach. A test battery of 13 ability tests and 3 standardized achievement tests was given to a sample of 1224 13-year-olds, and a hierarchical model with 13 latent ability variables was fitted to the covariance matrix. No gender differences were found in the structure of cognitive abilities. There were mean differences in favor of females in general intelligence and in the broad crystallized intelligence factor. Males had higher means on most spatial dimensions, for which they also were more variable. Males also had higher means on narrow numerical and verbal ability dimensions.

A correction model for differences in the sample compositions: the degree of comparability as a function of age and schooling

BackgroundSince the early days of international large-scale assessments, an overarching aim has been to use the world as an educational laboratory so countries can learn from one another and develop educational systems further. Cross-sectional comparisons across countries as well as trend studies derive from the assumption that there are comparable groups of students in the respective samples. But neither age-based nor grade-based sampling strategies can achieve balanced samples in terms of both age and schooling. How should such differences in the sample compositions be dealt with?MethodsWe discuss the comparability of the samples as a function of differences in terms of age and schooling. To improve the comparability of such samples, we developed a correction model that adjusts country scores, which we evaluate here with data from different IEA (International Association for the Evaluation of Educational Achievement) studies on reading at the end of primary school.ResultsOur study demonstrates that ignoring differences in age and schooling confounds league tables and hides actual trends. In other words, cross-sectional comparisons across countries as well as trends within countries are affected by differences in the sample composition. The correction model adjusts for such differences and increases the comparability across countries and studies.ConclusionsResearchers who use the data from international comparative studies for secondary analyses should be aware of the limited comparability of the samples. The proposed correction model provides a simple approach to improve comparability and makes the complex information from international comparisons more accessible.

Changes in the Multi‐Level Effects of Socio‐Economic Status on Reading Achievement in Sweden in 1991 and 2001

This study examines the changes in educational inequality at the school‐ and individual‐levels in 1991 and 2001. Comparisons are made between the IEA Reading Literacy Study 1991 and the so called 10‐Year Trend Study in PIRLS 2001. The between‐school differences in reading achievement variance and the size of the relationship between SES and reading achievement at school‐level are the two main focuses. Applying a two‐level structural equation modeling technique, it was found that the between‐school differences in reading achievement were greater in RLS 2001 than in RLS 1991, and the school SES effect has been strengthened over time. These changes indicate that educational inequality in Sweden has increased over time. Some possible explanations to these changes are discussed with reference to school reforms and demographic changes.

The dimensional structure of reading assessment tasks in the IEA reading literacy study 1991 and the Progress in International Reading Literacy Study 2001

The study investigates measurement properties of the reading assessment tasks used in the IEA 1991 Reading Literacy (RL) study and the Progress in International Readings Study (PIRLS) 2001 study. The analysis is based on data from the Swedish PIRLS study, comprising 16,676 students in grades 3 and 4. As an extension to the basic design, not only the PIRLS tasks were administered, but each student also completed 1 of the 2 booklets from the 1991 study. Using missing-data modeling techniques, confirmatory factor analysis models were estimated and tested for the complete set of reading tasks. Results show that both sets of tasks measure comprehension of reading continuous text, but that each of them also represents unique sources of variance. In the RL instrument, reading speed is one such component, and performance on tasks from the documents domain is another. The PIRLS tasks are influenced by the requirement to produce constructed responses. It is also demonstrated that the shared context among items referring to a particular text is a source of systematic variance.

Northern Lights on TIMSS and PIRLS 2011 : Differences and similarities in the Nordic countries

How is reading literacy taught in Nordic classrooms, and how is this influenced by the curricula? How can we improve mathematics teaching in Nordic classrooms? What is the relationship between scho ...

Teachers and tests: assessing pupils' reading achievement in primary schools

The purpose of the present study was to examine validity aspects of teachers’ judgements of pupils’ reading skills. Data come from Swedens participation in the Progress in International Reading Literacy Study (PIRLS) 2001, for Grades 3 and 4. For pupils at the same achievement levels, as measured by PIRLS 2001 test, teachers’ judgements of pupils’ achievement levels varied from one teacher to another. Moreover, there were significant differences between teachers’ judgements in Grades 3 and 4. Teachers in Grade 3 who had taught their pupils for almost 3 years showed higher correspondence between their judgements and pupil achievement within classrooms than 4th-grade teachers who, typically, had only taught their pupils for approximately 1 semester at the time of the data collection. The results indicate that teachers’ judgements and tests can be useful within classrooms, but that teachers may need external assessments to calibrate judgements over classrooms.

Measures of self‐reported reading resources, attitudes and activities based on latent variable modeling

In the IEA PIRLS International Report, a number of indices of reading‐related constructs were introduced for the purpose of explaining the variation in reading achievement. These indices, however, raise issues of reliability and validity, due to the way in which they are derived and due to the quality of the data. Targeting these issues, the current study investigates the measurement properties of some of these reading‐related factors by a multivariate latent variable modeling approach. The data is drawn from the PIRLS questionnaires of the fourth graders in six countries. On the basis of a series of confirmatory factor analytic models individual factor scores of these constructs are estimated. The results indicate that the most significant advantage of the suggested approach is that it makes efficient use of the available data to estimate factor scores with higher reliability and validity than the observed index scores.

Valid inferences of teachers’ judgements of pupils’ reading literacy: does formal teacher competence matter?

The primary aim of this study was to investigate whether and, if so, how formal teacher competence relates to the relationship between pupil reading achievement and teachers judgements’ of pupils’ reading achievement. The data come from the Swedish participation in the Progress in International Reading Literacy Study (PIRLS) 2001 in Grade 3. Information was obtained from pupils (N = 5271) and teachers (N = 351). Analyses were conducted using multilevel structural equation modeling (SEM) with random slopes. Teacher competence was operationalized using multiple observed indicators and defined as a latent variable. A higher correspondence between teacher judgements and pupil reading achievement for teachers with higher competence was found. The results highlight the importance of teacher competence in assessment practice.

Linking Large-Scale Reading Assessments: Measuring International Trends Over 40 Years

ABSTRACT Since the start of the new millennium, international comparative large-scale studies have become one of the most well-known areas in the field of education. However, the International Association for the Evaluation of Educational Achievement (IEA) has already been conducting international comparative studies for about half a century. The present study aims to demonstrate how to link recent and older studies onto the same scale in order to study long-term trends within and across countries. It discusses the comparability of the assessment material in the Progress in International Reading Literacy Study (PIRLS) and previous IEA studies on reading at the end of primary school. Thereafter, we use a concurrent calibration of all item parameters to link the studies onto a common IRT scale extending from 1970 until the present.

Gender Differences in Hierarchically Ordered Ability Dimensions: The Impact of Missing Data.

In this study, the impact of missing data on estimates of gender differences in hierarchically ordered ability dimensions is investigated. The data consist of 13 ability tests on which the whole sample of 1,224 13‐year‐old students has information and 3 standardized achievement tests on which a reduced sample of 981 participants has information. Utilizing missing data techniques for latent variable models, the study also becomes a validation of previously reported gender differences in latent hierarchical ability dimensions. In the previous analysis, the 243 students lacking data were assumed to be missing at random and thus excluded from the analysis. The attrition was found to have an impact on both the hierarchical model and on gender differences in latent dimensions. The attrition appeared biased with respect to general achievement and gender. When the cases with missing data were included in the analysis, the structure of the model remained stable and strengthened in some respects. The female advanta...