The effects of acceleration on students' achievement in senior secondary school mathematics: a multilevel modelling approach

Despite continuing curiosity about the effects of content acceleration on student performance there is little quantitative confirmation of the benefits of these programs. This research attempts to fill this gap considering four years of Victorian Curriculum and Assessment Authority (VCAA) data relating to achievement in mathematics. These secondary data constitute experimental data for content acceleration in that schools without acceleration programs provide control data. However, the acceleration decision is not taken randomly by schools so these data are only quasi-experimental in nature. The measures used for mathematical performance (Mathematical Methods and Specialist Mathematics scores) are accepted as reliable and valid by the Victorian education system. Controlling for individual characteristics such as gender and prior knowledge, and allowing for moderation effects due to school setting and school sector, the effects of content acceleration are estimated using multilevel modelling. The results suggest that content acceleration is beneficial, especially for students with higher prior knowledge scores. The quasi-experimental nature of the data means that only a tentatively causal relationship between acceleration and mathematical performance can be claimed. However, the statistical control of other factors means that this conclusion can be generalised to other states, other countries and, probably, to other subjects.