This paper is part of the Lärom project exploring the effects of restructuring of teacher education in Denmark and Sweden. As shown by Börjesson (2003 ) teacher education might be characterized as belonging to a dominated field within tertiary university education. In his paper at this conference, Hugo Wikstrøm(2004)has created four groups from the Lärom material under the main headings: "indistinct, professional, progressive and student-centred", acknowledging that in many respects the students seem to be quite homogeneous. This paper tries to explore in further detail the internal differences there nevertheless are in such a large material and the reasons for them. Differences might vary according to a huge number of factors, for this reason the present paper has concentrated on an analysis of the Danish case only, and tries to trace variations in the data which might be attributed to gender and social background. Other variations due to e.g. age, type of programme as well as country have been left for later analysis. The Lärom contribution by Gill and Johnson discusses regional variances. For the present purpose several different statistical techniques have been used, but first and foremost we have tried to use the correspondence analysis, known from Bourdieu´s impressive study "Distinction"(Bourdieu 1979/1986). This analysis has been supplemented by more common correlation studies. One of the problems in using correspondence analysis with attitudinal questions is that the questions might be interpreted differently by different people with otherwise antagonistic backgrounds and circumstances, so we do not claim to have found out the final truth about certain groups of teacher students, instead we hope that the results will give the opportunity for critical reflections and informed discussions. For excellent technical assistance and a lot of patience I thank Sverker Lundin, Dept. of Education, Uppsala University. This article is one of my contributions to the Swedish-Danish research project Lärom. All technical details on the data and the publications of the project can be found in the different electronic and paper publications referred to in the bibliography. Correspondence analysis is briefly presented in Lundin 2004