Examining School Leadership Practices in Multigrade and Non-Multigrade Schools: Evidence from Latin America and the Caribbean

• In Event: Inclusive Social Change in Latin American Education

Proposal

Background & Framework
In 2004, one-third of schools worldwide were multigrade, mostly located in high-poverty rural areas, and the number has increased in recent years (Cornish & Taole, 2021; UNESCO, 2004). In the Global South, multigrade schools reach students in the most remote areas and are characterized by a lack of resources and support from educational authorities. Despite the prevalence of these schools, scholars and policy-makers have focused their efforts on studying and implementing policies and practices relevant to mono-grade schools.
School leadership is a key factor to achieve educational quality. There is a vast body of evidence that shows that effective school leadership has a positive impact on student’s achievement (Brown, 2010; Grissom et al., 2021; Jackson, 2000; Klein et al., 2018; Phillips et al., 2022; Robinson et al., 2008; Ross, 2022; Spillane et al., 2001, 2004). In this context, some authors have developed a set of school leadership practices linked to an increase in student achievement (Bryk et al., 2010; Halverson & Kelley, 2017; Robinson et al., 2008; Robinson & Hargreaves, 2011). These frameworks have many common components: focus on teacher quality and development, ensuring a nurturing and learning-centered environment, and working together with parents/community. However, it is not possible to implement these practices without having leaders dedicated exclusively to a principal’s role, something that does not happen in the case of multigrade schools.
Despite the existing evidence on the importance of school leadership for educational success globally, studies in the Global South have not examined the impact of school leadership on student’s achievement in multigrade settings. For that reason, this quantitative paper seeks to answer the following questions: 1. Do leadership practices differ between multigrade and non-multigrade schools? If so, how? 2. What is the relationship between school leadership and students’ achievement? 3. Is the relationship between school leadership and students’ achievement different in multigrade vs non-multigrade schools?
Comparing multigrade and non-multigrade schools, this paper will help generate evidence on the relevance of school leadership practices in multigrade settings, helping to understand how school leadership could work in different types of schools. Additionally, this paper will contribute to expand our understanding of school leadership by building upon existing theories and frameworks and exploring their applicability in rural multigrade school settings. The expected theoretical contribution is to generate new knowledge on how school leadership operates in these contexts and identifying areas where existing theories may need to be adapted or extended.

Data & Methods
We use 2019’s ERCE data, which is a large-scale study conducted by UNESCO's Latin America and Caribbean Regional Office (OREALC) that aims to assess the quality of education in 16 countries in the region. We focus on the principal’s questionnaire and also include school level scores in math and language for 3rd and 6th grade students.
To assess research question 1 (RQ1), we run a t-test to compare leadership scores between schools according to multigrade status. Then, we disaggregate the leadership index and compare its four components separately: how often principals (i) observed how teaching is done, (ii) created opportunities for teachers to collaborate with each other, (iii) took actions to favor the improvement of teachers’ pedagogical skills, and (iv) provided parents or guardians with information about school and student performance.
To assess research question 2 (RQ2), we run a multiple regression with clustered standard errors at the school level. Our main dependent variables are school averages for math and language scores. Our main explanatory variable is the leadership index provided by ERCE. Control variables include whether school is public, in a rural setting, the principal’s gender and years of experience, and grade.
Finally, to assess research question 3 (RQ3), we run the same model but separately for multigrade and non-multigrade schools. Additionally, we run a full interaction model, in which each explanatory variable is interacted with multigrade status, to assess whether difference in coefficients for each regression is statistically significant.

Findings
For RQ1, we find that self-reported leadership practices are higher for principals in non-multigrade schools (b= 0.012) compared to principals in multigrade schools (b= -0.073), and this difference is statistically significant (p<0.05). Furthermore, leadership practices look different in these schools. In non multigrade schools, the most commonly reported leadership practice is taking actions to favor the improvement of teachers’ pedagogical skills. On the other hand, for multigrade schools, the most commonly reported practice is providing parents or guardians with information about student performance.
In our analysis of RQ2, we find that a baseline regression with no controls shows a positive and significant relationship between leadership and both math and language scores at the school level (b=0.04 and b=0.05). However, when we include other controls, these overall relationships are no longer significant.
For RQ3, results indicate that the principal's leadership practices are positively related to school level performance in both math (b=0.10, p<0.01) and language (b=0.05, p<0.05) in multigrade schools. This is not the case for non-multigrade schools, where the coefficient for leadership was not significant.. Additionally, in both cases, the full interaction model suggests that the relationship between leadership and school performance is indeed different between multigrade and non-multigrade schools.

Conclusion
This study reveals important differences in school leadership practices and their impacts between multigrade and non-multigrade schools in Latin America and the Caribbean. While principals in non-multigrade schools report higher overall leadership practices, the nature of these practices differs between school types. Notably, leadership practices show a significant positive relationship with school-level performance in math and language specifically in multigrade schools, a relationship not observed in non-multigrade schools. These findings highlight the unique importance of school leadership in multigrade settings, suggesting that existing leadership frameworks may need adaptation for these contexts. Furthermore, the study underscores the need for tailored approaches to school leadership in multigrade schools, which are prevalent in rural and high-poverty areas of the Global South. Future research and policy initiatives should consider these differences to better support and enhance leadership in diverse school settings, potentially improving educational outcomes in underserved areas.

Authors

• Micaela Wensjoe, The University of Wisconsin-Madison

• Carla Zenobia Glave Barrantes, The University of Wisconsin-Madison