Methods Talk - Population Research Center

Date: 04/26/2022

Abstract

Beyond exclusion: The role of high-stake testing on attendance on the day of the test

Magdalena Bennett (University of Texas at Austin), Christopher Neilson (Princeton University), & Nicolás Rojas (Columbia University)

High-stake testing plays a crucial role in many educational systems, guiding policies of accountability, resource allocation, and even school choice. However, non-representative patterns of attendance can skew how useful these measures are for accomplishing their main objective. Are we really measuring the quality or performance of a school if a non-representative sample of their students is taking the test?

In this paper, we study the effect of high-stakes testing on student composition of attendance on the day of the test using rich administrative data from Chile and daily attendance. By combining an event-study framework and a machine learning prediction approach we focus on three main objectives: (i) understand the average effect of high-stakes tests on school attendance across grades and performance, (i) help improve current imputation methods, and (ii) identify schools that could potentially be incentivizing non-representative patterns of attendance.

Our analysis reveals that the effect of high-stakes testing on attendance is highly heterogeneous across grades, and younger students are more affected than the rest. Most notably, the increased attendance of higher-performing students (measured according to their GPA) seems to be crucial to selective patterns of attendance. These results are robust to other explanations, such as students’ own self-selection or exemptions for students with disabilities.

Some preliminary results also show high levels of heterogeneity between schools regarding their observed and predicted patterns of attendance. While some schools promote attendance for higher performance students, others present clear patterns that are consistent with discouraging lower-performing students from attending. These findings are particularly relevant in a context where missing scores are imputed using a general blanket policy.

Download code 💻 Open slides 📂