My four-year PhD journey has come to an end and made me realize that this would not have been possible without the tremendous support and encouragement from so many people. First, my most sincere gratitude goes to my main supervisor professor, Rolf Vegar Olsen. With a great balance between giving me the freedom and responsibility to shape my projects, and the continued support and guidance to reach my (sometimes far-fetched) goals, this has been an interesting, fun, challenging, and personally and professionally enriching journey. You have always provided academic guidance, and your support, encouragement, and constructive feedback on my work have been extremely valuable to me. I would also like to give my deepest thanks to my co-supervisor, professor Ove Edvard Hatlevik, whom I got halfway through. Your support, valuable comments, and engaging discussions have been immensely important to me. I am also grateful to my second co-supervisor, professor Ola Erstad, who introduced me to the core of the field and has offered his support and useful comments at various stages of my work. My heartfelt thanks go to professor Mark Wilson at the Berkeley Evaluation and Assessment Research Center at the University of California, Berkeley. Thank you so much for supporting my work and generously providing resources and collaborators for all my visits to Berkeley. I feel extremely lucky to have had the chance to work on such a novel assessment and to have been guided by you. I also want to thank David Torres Irribarra for applying for the Peder Saether Center grant with me. Daniel Stansfield, thank you so much for the initial introduction to the technicalities of the assessment software. I am infinitely grateful to my co-author and friend Ronny Scherer. You are indeed a great researcher, but for me also a role model in many ways. Thank you for increasing the quality of my work, for making me sweat and work hard, and for the many great laughs which made all the efforts worth it. Also a huge thanks goes to my co-authors Jo Tondeur, Inger Throndsen, and Perman Gochyeev for the collaboration, motivation, help, and discussions. Your support and generosity have meant a lot to me, and I feel fortunate to have had all of you as co-authors. I need to go a little back in time and thank

This chapter examines how crucial input and process characteristics of schooling are related to cognitive student outcomes. It was hypothesized that teacher quality predicts instructional quality and student achievement, and that instructional quality in turn predicts student achievement. The strengths of these relations may vary across countries, making it impossible to draw universal conclusions. However, similar relational patterns could be evident within regions of the world. These hypotheses were investigated by applying multi-level structural equation modeling to grade four student and teacher data from TIMSS 2011. The sample included 205,515 students from 47 countries nested in 10,059 classrooms. Results revealed that teacher quality was significantly related to instructional quality and student achievement, whereas student achievement was not well predicted by instructional quality. Certain characteristics were more strongly related to each other in some world regions than in others, indicating regional patterns. Participation in professional development activities and teachers' sense of preparedness were, on average, the strongest predictors of instructional quality across all countries. Professional development was of particular relevance in Europe and Western Asian/Arabian countries, whereas preparedness played an important role in instructional quality in South-East Asia and Latin America. The ISCED level of teacher education was on average the strongest predictor of student achievement across all countries; this characteristic mattered most in the Western Asia/Arabia region.

BackgroundQuestionnaires are commonly used to collect patient, or user, experiences with health care encounters; however, their adaption to specific target groups limits comparison between groups. We present the construction of a generic questionnaire (maximum of ten questions) for user evaluation across a range of health care services.MethodsBased on previous testing of six group-specific questionnaires, we first constructed a generic questionnaire with 23 items related to user experiences. All questions included a "not applicable" response option, as well as a follow-up question about the item's importance. Nine user groups from one health trust were surveyed. Seven groups received questionnaires by mail and two by personal distribution. Selection of core questions was based on three criteria: applicability (proportion "not applicable"), importance (mean scores on follow-up questions), and comprehensiveness (content coverage, maximum two items per dimension).Results1324 questionnaires were returned providing subsample sizes ranging from 52 to 323. Ten questions were excluded because the proportion of "not applicable" responses exceeded 20% in at least one user group. The number of remaining items was reduced to ten by applying the two other criteria. The final short questionnaire included items on outcome (2), clinician services (2), user involvement (2), incorrect treatment (1), information (1), organisation (1), and accessibility (1).ConclusionThe Generic Short Patient Experiences Questionnaire (GS-PEQ) is a short, generic set of questions on user experiences with specialist health care that covers important topics for a range of groups. It can be used alone or with other instruments in quality assessment or in research. The psychometric properties and the relevance of the GS-PEQ in other health care settings and countries need further evaluation.

No abstract

The Programme for International Student Assessment in 2006 included several measures of students' interest in science. These measures were constructed by combining information from several items where students are asked to respond to statements along Likert scale categories. Since there is evidence for Likert scales providing culturally biased country scores, we demonstrate in this article that the relative profiles of interest can be meaningfully analysed across countries. Hence, we have developed national relative profiles of interest in science constructed from the country-and item-specific residuals at the item level. Subsequently, these relative profiles of interest have been used as input in a cluster analysis providing identification of distinct groups of countries with similar item-by-item patterns of interest in science. The most notable feature of the analysis is an overall division between two larger groups of countries, roughly corresponding to European/Western countries in one group and non-European countries, with only a few exceptions, in the other group. A number of meaningful clusters of countries, partly defined by language and partly by localisation, are identified within each of the two main clusters. In order to develop a more detailed understanding of the characteristic features of the various clusters, descriptive information about the items is included in the analysis. The most notable finding is the strong relative preference for life and health issues among the non-European countries, contrasted with the distinct favouring of items relating to physical/technological systems in the European/Western countries.

No abstract

BackgroundThe Psychiatric Out-Patient Experiences Questionnaire (POPEQ) is an 11-item core measure of psychiatric out-patients experiences of the perceived outcome of the treatment, the quality of interaction with the clinician, and the quality of information provision. The POPEQ was found to have evidence for reliability and validity following the application of classical test theory but has not previously been assessed by Rasch analysis.MethodsTwo national postal surveys of psychiatric outpatients took place in Norway in 2004 and 2007. The performance of the POPEQ, including item functioning and differential item functioning, was assessed by Rasch analysis. Principal component analysis of item residuals was used to assess the presence of subdimensions.Results6,677 (43.3%) and 11,085 (35.2%) psychiatric out patients responded to the questionnaire in 2004 and 2007, respectively. All items in the scale were retained after the Rasch analysis. The resulting scale had reasonably good fit to the Rasch model. The items performed the same for the two survey years and there was no differential item functioning relating to patient characteristics. Principal component analysis of the residuals confirmed that the measure to a high degree is unidimensional. However, the data also reflects three potential subscales, each relating to one of the three included aspects of health care.ConclusionsThe POPEQ had excellent psychometric properties and Rasch analysis further supported the construct validity of the scale by also identifying the three subdimensions originally included as components in the instrument development. The 11-item instrument is recommended in future research on psychiatric out-patient experiences. Future development may lead to the construction of more precise measures of the three subdomains that the POPEQ is based on.

The thesis presents secondary analyses of data from a large-scale international assessment of students' achievement in mathematics, reading and science entitled the Programme for International Student Assessment (PISA) which is initiated and monitored by the OECD. The thesis has been conducted and supervised at the Department of Teacher Education and School Development at the Faculty of Education, University of Oslo. The work presented in the dissertation is framed by two overarching questions relating specifically to the science achievement data in PISA: What is the nature of the information in the cognitive science items in PISA?, and; what is the analytical potential of this information? These two questions have been used as guides in several explorative analyses of the science achievement data in PISA presented in three published papers. The background for posing these general questions is the fact that when the achievement data are reported from these studies all the item specific information has been lost: Initially students' responses to each of the items have been marked by a number of specific codes capturing qualitative aspects of the responses. These codes are subsequently reduced or aggregated into credits or score points. Then the item scores are aggregated into one (or a few) reliable and valid overall achievement scores. The first paper is primarily an introduction into this general background and a rationale is developed for why the study of the single items before they are aggregated is relevant and possible. The second paper uses homogeneity analysis in the study of the codes used in the initial marking before they are reduced to score points. The third paper studies the relative achievement profile across all items and across all participating countries by applying cluster analysis. The analysis demonstrates that countries are clustered and each cluster has a distinct profile across the items, and these profiles are not captured in the overall achievement scores. One overall message in the thesis is that large-scale international comparative achievement studies do not only provide high-quality data of students' overall science achievement across the world. There is also a fine-structure in the data across the single items. It is envisaged that these data may also be used in secondary analyses targeting more specific research questions within the research field of science education across the world.