摘要：This reply responds to a comment by Cannon (2011) that opens the debate on consistency of the effectof downloading PowerPoint slides before lectures on students’ exam performance. Cannon (2011)points out potential endogeneity problems in Chen and Lin (2008) and attempts to explore theunconditional mean effect of downloading PowerPoint slides for the full sample. In this reply, we firstlyargue that the estimates in our original article are consistent since the effect of interest is the“conditional” treatment effect but not the unconditional mean effect. We provide explanations for ourrationale of estimating the “conditional” treatment effect. Secondly, we propose a modifieddownloading variable to replicate Cannon’s analysis. Our results suggest that downloading PowerPointslides before the exam does not produce a significant effect on absent students’ exam performancewhich is different from the results in Cannon (2011). Our analysis does support Cannon’s argument thatstudents fixed effects are different across different attendance status.
