Grade inflation in Slovenia Darko Zupanc and Matevž Bren
Summary: Grade inflation is defined as an upward shift of average school grades over an extended period of time without a corresponding increase in student achievement. This issue was pointed out years ago when experts in the U.S. became aware of the year–to-year rise in the grade point average (GPA) and also when a similar situation occurred with the internal grading of various subjects of general Matura in Slovenia. This article shows differences (academic years 87/88 and 05/06) in the distribution of overall achievement in primary schools in Slovenia and in high graded internal parts of the general Matura exams over time (1995 to 2008). The study confirms the prognosis given nine years ago that predicted a yearly increase in internal grades in the school-based parts of the Matura exams (i.e., oral exam, laboratory work, coursework, etc.); these are now approaching a 90% average, and in some subjects even exceed that percentage. The study analyzes the internal grades of subject groups with the same proportion (20%, 25%, 30%, and 40%) of the internal part of the exam. Internal grades are higher in groups with a higher proportion of the internal exam; the fastest growing internal grade average is in the group with 30% of internal proportion. The second highest slope of increase is in the groups with 40% and 25% of the internal part, where the group with the 40% proportion of school-based part starts much higher in the beginning year (2000). The minimum annual growth has been recorded in the group with the lowest, i.e. 20%, internal part. All upward trends are statistically significant. Grade inflation, meaning that higher grades do not derive from a corresponding increase in student achievement, is confirmed by correlation research between the internal and the external part of the Matura exams (correlations are relatively low), and by the Matura Subject Commission’s trial, where teachers’ and external examiners’ grades regarding the same objectives were compared.