Summary and discussion

The aim of this study was to describe and analyze a student population with reference to personality traits, mental health and social conditions, and the correlation of these factors with study situation and study results. The possibility of predicting academic performance during the first years at the university was investigated. The methods used were clinical and epidemiological. From first-year students studying four different main subjects at the University of Uppsala 1968, a sample of 200 students was chosen. During their first semester, 195 of the students could be interviewed. After two years 186 students could be re-interviewed. At the first interview social background conditions, physical and mental health during childhood and adolescence, school conditions and school reports and motivation for university studies were dealt with. An assessment of the students’ personality dimensions, according to Sjobring’s model, was made by the investigator. Self-rating, by means of two personality inventories, often used in Sweden, namely MarkeNyman’s temperament scale (MNT) and Eysenck’s Personality Inventory (EPI) was done by the students. At the second interview, difficulties experienced by the students during their first two years of studies were discussed, and their physical and mental health during this period was recorded. A new assessment of personality traits was made according to the same methods used at the initial interview. Study results, shown as points for the initially chosen subject after one and two years respectively, and total study results (all subjects studied) after two years were recorded. For analyzing the correlation between certain explanatory variables and study results as dependent variables, multiple regression analysis and Automatic Interaction Detector (AID) analysis were used. For the assessment of social background conditions, motivation for university studies, presumed adaptability to studies, contact with comrades, attitude to studies, physical and mental health, rating scales of a ranking type were constructed. The validity of such scales is, of course limited, and they give only an approximation of the required measurements. The reliability of the scales was not tested. The scales based on clinical experience, naturally included sets of interdependent variables. Nevertheless, the scales may supply more information than is obtainable by the description of isolated phenomena.