Abstract
This investigation examined the performance of students (9–12 year olds) over a three-year period as they solved graphics-based mathematics items which are commonly found in standardised tests. There were statistically significant improvements in the performance of students (across each of the six language types) in each year of the study. Mean scores for boys were higher than girls on most measures with statistically significant differences in Axis (number line), Map and Retinal-list categories.