 |
Recent research shows student effort in the ‘production function' of test scores is sub-optimal (Levitt et al, 2012), and that provision of incentives to enhance academic performance could be a potentially cost-effective strategy. However, not much is known about the ‘optimal' structure of incentives, and whether incentives work under real test scenarios. In this paper, we examine the relative efficacy of a battery of individual and group-based incentives during NAPLAN tests at a high school in Queensland. The results suggest gains in performance across all incentive treatments for Year 7 students, and for some treatments for Year 9 students. Among all treatment, we find group-based incentives was most effective for Year 7 students, which suggests a strong role of ‘social incentives'. Furthermore, we find heterogeneity of gains in test scores across ability distribution – the high-ability students in Year 7 and low-ability students in Year 9 demonstrated highest gains under the group-based incentives. The insights from this field experiment are not only important from an educational policy perspective, but have wider implications for any organisation trying to increase effort-levels of its employees.
