School Leaders on Safety and Learning Goals

How to cite

Sheth, Ketki; Pugatch, Todd. 2025. "School Leaders on Safety and Learning Goals." AEA RCT Registry. AEARCTR-0017297. https://doi.org/10.1257/rct.17297-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Ketki Sheth* , Todd Pugatch

* Primary investigator

Registry ID
AEARCTR-0017297
DOI
https://doi.org/10.1257/rct.17297-1.1
First registered
Last updated
Status
Completed
Trial dates
2025-11-17 – 2025-12-15
Countries
Uganda ()
JEL codes
I21, I30
Registry record
http://www.socialscienceregistry.org/trials/17297

Abstract

Across education systems, school leaders play a central role in shaping two of the most fundamental dimensions of student outcomes: learning and safety. Students cannot learn effectively in environments where they feel unsafe, and even the best-managed classrooms will fail to reach their potential if teaching and learning are weak. Effective leadership therefore requires not only improving academic performance but also maintaining a positive, safe school culture. Yet, while both are widely recognized as core to school quality, there is limited evidence on how leaders prioritize and act on these responsibilities in practice.

This study examines how secondary school leaders in Uganda navigate these two domains of leadership—student performance and student safety. On student performance, we explore how leaders engage with data from the Uganda Certificate of Education (UCE) exams—one of the key sources of performance feedback available to schools. We ask three related questions:

(1) How do school leaders prioritize improvement across high- and low-achieving students and different subjects?
(2) How well do school leaders know the skills of their students?
(3) How well do school leaders understand and use performance data when setting strategies for student achievement? Do exam results provide genuinely new or actionable information?

Turning to school safety, we examine how leaders respond to scenarios of teacher and student misconduct—an equally critical but often less visible dimension of school quality. In particular, we investigate whether the gender of the student victim influences leaders’ perceptions of severity and their proposed disciplinary actions.

Intervention

We study two randomized interventions: 1) sharing student learning data in an aggregated format, and 2) randomizing gender (and severity) on vignette questions around teacher misconduct. We also include a discrete choice experiment in which we randomize four attributes around school performance.

Experimental design

This is a survey and lab in the field experiment with school leaders in Uganda, in which we have randomized key elements (described above).

Design details

Randomization method
randomization done in survey software (surveyCTO)
Randomization unit
Individuals (i.e., school leaders)
Clusters
Between 200 and 350 school leaders.
Observations
Between 200 and 350 school leaders.
Arms
Each random assignment has a 50% probability (i.e., 100 to 175 would be randomly assigned to a given treatment).
Primary outcomes
For the first randomization on student performance data, our key outcomes are correctly assessing student performance, how school leaders prioritize and identify subjects, and the action plan they would take in response to performance data. For the second randomization on school-based teacher misconduct, we explore whether school leaders respond more to misconduct based on student gender. For the discrete choice experiment, we analyze how school leaders value trade-offs between performance improvements at different points in the distribution of exam results.
Sponsors
Name: Center for Effective Global Action Location: Berkeley, California Url: https://cega.berkeley.edu/
Partners
Name: Promoting Equality in African Schools Partner_type: ngo Url: https://www.peas.org.uk/

Cited by

No citation data available for this registration yet.

Keywords