Table 1.
Characteristics of the control variables.
Fig 1.
Distribution of test scores and teachers’ grades by students’ weight category.
The upper-left graph shows Kernel density estimates of the distribution of test scores in German, the upper-right graph in mathematics. The lower-left graph shows the distribution of teachers’ grades by students’ weight category in German, and the lower-right graph in mathematics.
Fig 2.
Teachers’ grading bias (measured in log-odds) according to students’ BMI.
The graph shows log-odds ratios from the 3-level hierarchical ordered logit model on mathematics grades (left) and German (right). Bars represent 95% confidence intervals. Model 1 adjusts for students’ competence level and socio-demographic variables, while model 2 also includes students’ psychological traits, and school-related attitudes and behavior. N = 3,754. Analysis conducted on 50 multiply imputed datasets.
Fig 3.
Teachers’ grading bias (measured in average partial effects) according to students’ BMI.
The graph shows average partial effects from the 3-level hierarchical ordered logit model on mathematics grades (upper part) and German (lower part). Bars represent 95% confidence intervals. The models are adjusted for students’ competence level, socio-demographic characteristics, psychological traits, and school-related attitudes and behavior. N = 3,754. Analysis conducted on 50 multiply imputed datasets.
Table 2.
Teachers’ grading bias (measured in average partial effects) over the grades distribution, according to students’ BMI and gender.
Fig 4.
Predicted probabilities of obtaining a low grade by BMI and gender.
Predicted probabilities (and 95% confidence intervals) derived from hierarchical ordinal logistic regression models of teachers’ grading in German (left graph) and mathematics (right graph), adjusted for students’ competence level, socio-demographic characteristics, psychological traits, and school-related attitudes and behavior. N = 3,754. Analysis conducted on 50 multiply imputed datasets.