Due to societal changes there is a growing need for distant and adult learning. The reason to participate in education and the choices that students make may differ. In this study the factors age, gender, rate of studies and parenthood have been analysed in order to see how these relate to different motivational factors for choosing a web-based course. The data has been based on a questionnaire, covering 1270 beginner students in the spring semester of 2011 and contains their background characteristics and items focusing on their motives. These could be categorized into four different motives: (1) Format, (2) Content, (3) Economic, and (4) Curiosity. The results showed that Format was regarded as the most important factor for choosing an Internet-based course, followed by Content, Curiosity and the Economic factor. Furthermore, group differences were investigated with respect to age, gender, parenthood and rate of study. The findings show that distant education fulfils an important function for mature students, women and students with children. These groups presumably consider the flexibility that web-based courses provide advantageous. Family situations or working-life obligations may contribute to this. Changes in people’s working lives are likely to continue, which presumably increases the demand for flexible learning situations.

This study examined student police officers’ job values and differences in job values with respect to gender and educational background. Three hundred fifty-two Swedish first-year police students responded to a job value questionnaire. Psychometric analyses indicated that there are three dimensions of job values: intrinsic, altruistic, and extrinsic. In general, students rated altruistic values as more important than intrinsic/ leisure and extrinsic values. Females valued intrinsic and altruistic aspects of a job as more important than males did. In conclusion, the three dimensions of job values, as well as the differences found between males and females are supported by previous research on job values. Future studies should be directed on examining whether police students’ job values change during education and in the transition to work.

Test takers’ perceptions of fairness in testing have rarely been studied. The aim of this study was to develop, apply and validate scores from an instrument measuring fairness perceptions in higher education. The theoretical rationale for examining test takers’ perceptions was the model of organizational justice, which consists of aspects of procedural- and distributive justice. A questionnaire measuring perceptions of procedural justice was administrated to 441 test takers of the Swedish Scholastic Assessment Test. Although the results indicated adequate validity evidence for the procedural justice scale, there is room for improvement of the overall structural validity on this measure. Finally, the test takers’ perception of principles of distributive justice is discussed in relation to higher education selection.

This thesis consists of four papers, all of which are concerned with the validation of selection to higher education. In paper [I] student performance was observed for a period of three years with a sample of three different admission groups. The purpose of the study was to see if there were any differences in academic performance between the three groups. The main result was that students admitted on the basis of their SweSAT scores and credits for work experience were less successful than the other two groups. In paper [E] the concept of validity is presented and discussed. The paper presents a historical perspective in the changed meaning of the concept of validity. In the paper, validity is also discussed in relation to qualitative and quantitative approaches. Paper [IH] was an attempt to apply a multi-faceted concept of validity to the Swedish model for selection. The results indicated that the assumption about unidimensionality of the components can be questioned, i.e. the unidimensional model for the SweSAT, GPA and the criterion for students achievement showed poor model-fit. This paper also briefly discusses the values that can be connected to the constructs that the instruments intend to measure. The result can be regarded as one explanation for the difficult-to-interpret picture resulting from an analysis of the predictive validity for the SweSAT and the GPA. In conclusion, the results indicate that a validation of a selection procedure does not produce any conclusive answers. Instead it raises important questions and issues that need to be considered for decisions about the applied selection procedure. In paper [IV] the concepts of fairness and justice are discussed in relation to selection to higher education. One central question in a validation procedure is the purpose of the selection. When the selection is viewed from the perspective of principles for distribution, several different principles can be identified. The relation between justice/fairness and validity is discussed. When the selection procedure aims to fulfill different purposes, and when the consequences are contradictory, this leads to a validity problem. Seen together, these articles constitute not only an evaluation of the Swedish selection, but also a methodological discussion of validity and validation practice. Overall, Messick's (1989) theoretical framework for the analysis of validity is judged to be a powerful instrument for the validation of assessment systems.

Introduction: Sweden, together with several countries in the world, has no compulsory educational requirement prior to the drivers license examination. Having a valid license examination (i.e., a knowledge test and a driving test) is therefore important in order to ensure competent drivers. Method: The aim of this paper is to explore the possibility of a two stage testing approach for the Swedish licensure examination. In order to explore the relationship between the knowledge test and the road test, a descriptive and a logistic regression approach were carried out. Results: The result indicates that how the driving candidates perform on the road test depends on their performance on the knowledge test. Therefore, the Swedish drivers licensing examination is suitable for a two-stage testing model.