Measurement in Behavioral Sciences

Research at BVM spans a wide spectrum. By combining statistics, psychology, pedagogy, and cognitive science, the field contributes to ensuring that our assessments are fair, meaningful, and useful, while also providing important insights into how learning occurs. Researchers in the field develop tests, surveys, and assessment systems used in contexts such as education, the workplace, and healthcare. We also study cognitive aspects of memory and learning—factors that are crucial for promoting students' learning in school. Our research enables informed decision-making, evaluation of interventions, and the promotion of fairness and quality in education and public services. Research at BVM includes everything from the design of national and selection tests to how employers assess competence, or how cognitive and affective factors influence students' learning and performance. The research is also part of one of Umeå University's prioritized areas, focusing on human learning and the brain's plasticity throughout life. Below are some of our research areas.
Research Leader : Bert Jonsson

International Large-Scale Assessments

Research in this area focuses on the quality and comparability of large-scale, standardized knowledge assessments—both national and international. A central focus is ensuring that test results are fair and valid over time, across different test versions, and among various student groups. This includes analyses of how different forms of test administration, such as digital versus paper-based tests, affect results (known as mode effects), with particular attention to equity for individuals and groups. Several studies address international surveys such as PISA and TIMSS, where Swedish students' results are analyzed both separately and in comparison with other countries' education systems. The research also covers students' digital competence and its relation to performance in computer-based tests. An important methodological focus is psychometrics, especially measurement invariance—i.e., whether a test measures the same construct in the same way across different groups—as well as the significance of longitudinal analyses.
For more information, contact Associate Professor Hanna Eklöf or Associate Professor Ewa Rolfsman.

Admission and Selection for Higher Education

This research focuses on admission to higher education from a validity and fairness perspective. It is conducted in an interdisciplinary environment that combines expertise in pedagogy, psychometrics, and education economics. The studies address both selection models and whether the instruments used are reliable and relevant for predicting future academic success, as well as whether they can be considered fair from social and ethical perspectives. Key themes include the evaluation of traditional selection tools, such as grades and tests—with a particular focus on the Swedish Scholastic Aptitude Test (Högskoleprovet), which is developed at the department—and the development of alternative methods. The research also analyzes how factors such as gender, ethnicity, and socioeconomic background affect access to higher education. Both quantitative and qualitative methods are used, including statistical analyses of admissions data, policy reviews, and interview studies. The research environment collaborates with national authorities, universities, and international networks to contribute knowledge about fairness in the education system.
For more information, contact Associate Professor Christina Wikström.

Assessment and Grading in Schools

Through work with assessment support and national tests in mathematics and science subjects, extensive experience has been built around the analysis and interpretation of course and subject syllabi. This knowledge is applied in tasks and assessment guidelines aimed at supporting equitable grading in these subjects. Research in this area focuses, among other things, on the analysis of inter-rater reliability—that is, the extent to which different assessors make equivalent judgments of students' work. Additional studies address, for example, standard setting, the process used to determine the boundaries for different test grades at each new test occasion.
For more information, contact Project Manager Anna Lind Pantzare.

Test Development and Psychometrics

This research aims to develop and evaluate tests, surveys, and other measurement instruments used to understand people's knowledge, experiences, and health. It includes both education-related tools—such as national tests, the Swedish Scholastic Aptitude Test, and surveys on motivation, test anxiety, or mental well-being—and questionnaires used in healthcare to capture symptoms and functional ability. The work involves both the development of new instruments and the review and further development of existing ones. A central question is how well these tools measure what they are intended to (validity) and how reliable the results are over time and across different groups (reliability). To assess this, psychometric methods are used—statistical and mathematical techniques that analyze item functioning, relationships between responses, and the instruments' equivalence. These methods enable fair comparisons between versions, years, and groups, and include differential item functioning (DIF), classical test theory, item response theory (IRT), and multilevel analysis to study how student results correlate with background factors within and between education systems.
For more information, contact Associate Professor Per-Erik Lyrén.

Learning and Teaching

The thematic area Learning and Teaching studies how cognitive and affective factors influence students' learning and performance. Cognitive aspects include working memory, attention control, and impulse regulation, while affective factors involve self-perception, motivation, and anxiety. The research is based on longitudinal follow-ups, experiments, classroom studies, surveys, and large-scale data analyses. A long-term research line investigates test-based learning (retrieval practice) and Creative Mathematical Reasoning (CMR), in collaboration with UFM. The work includes both classroom interventions and brain imaging studies together with UFBI, focusing on the brain's role in learning. In recent years, research has focused on the relationship between math anxiety and performance. A parallel research line focuses on students' experiences of assessment and test situations. The group combines self-reports with digital process measures, such as response times, to map how affective factors influence performance. The work includes analyses of international databases such as TIMSS and PISA and has resulted in several internationally recognized publications.
For more information, contact Professor Bert Jonsson.