Das Dokument ist öffentlich zugänglich im Rahmen des deutschen Urheberrechts.

Using real effort to implement costly activities increases the likelihood that the motivations that drive effort provision in real life carry over to the laboratory. However, unobserved differences between subjects in the cost of real effort make quantitative prediction problematic. In this paper we present the slider task, which was designed by us to overcome the drawbacks of real effort tasks. The slider task allows the researcher to collect precise and repeated observations of effort provision from the same subjects in a short time frame. The resulting high-quality panel data allow sophisticated statistical analysis. We illustrate these advantages in two ways. First, we show how to use panel data from the slider task to improve precision by controlling for persistent unobserved heterogeneity. Second, we show how to estimate effort costs at the subject level by exploiting within-subject variation in incentives across repetitions of the slider task. We also provide z-Tree code and practical guidance to help researchers implement the slider task.