Abstract
Achievements of students in primary education have been the object of study over a long period. Both individual and contextual characteristics have proved their value for the prediction of these achievements. In this paper a theoretical model of the time and attention division of the teacher is elaborated. From this model, hypotheses are deduced about the labeling of a student as a problem student, and the possible referral of problem students to special education. Relative student characteristics are predicted to be of more importance than absolute characteristics. Hypotheses testing is based on data about 2,340 students in the Netherlands. The hypotheses are confirmed. Labeling of problem students is based on the relative achievements and relative behavior of students, while the referral to special education also can be predicted on the basis of relative achievements.