Learning for All: Insights on mid-adolescents’ literacy from a multi-site, multi-disciplinary, multi-method project

Group Submission Type: Formal Panel Session

Proposal

The proposed panel includes three studies that seek to understand factors associated with literacy achievement in mid-adolescents (grades 4-8; ages 9-14) across three countries: Botswana, Brazil, and Colombia. Instead of system-level policies or school-input measures, this panel examines students’ individual skills and relational dimensions of school learning that hold potential to augment learning by “chang[ing] children’s daily experiences” (Ganimian & Murnane, 2016). With the ultimate goal of understanding local and global levers for learning, we will present context-specific findings that emerged from applying a similar multiple-method and multi-disciplinary design across sites. Dr. Karen Mundy, globally recognized education specialist, will be our discussant.

Educational equity constitutes a pressing global challenge. Despite remarkable progress towards universal primary education, poor quality education is jeopardizing the futures of individuals and societies around the world (UNESCO, 2017). Given the widely documented and persistent literacy underperformance of adolescents in numerous countries and particularly the socioeconomically-based inequalities in literacy achievement (OECD, 2014; 2016), we focus on mid-adolescents’ literacy. Starting around 4th grade, reading comprehension becomes essential for learning across disciplines; yet, even after mastering basic early literacy skills (i.e., word recognition skills), large proportions of adolescents struggle to learn how to comprehend text. This is particularly worrisome today, when advanced literacy skills are more crucial than ever to navigate the educational, professional, and civic demands of information-based societies (Levy & Murnane, 2013).

Research in developing countries, however, has focused mostly on early literacy (Gove & Cvelich, 2011), with comparatively minimal investigation exploring mid-adolescents’ literacy. Whereas the Programme for International Student Assessment-PISA systematically documents 15-year-olds’ literacy achievement across countries (OECD, 2016), research on the earlier mid-adolescent years that lead to these outcomes is still scarce in developing countries. This gap is surprising given that developmental research unequivocally shows that strong early literacy skills are necessary yet far from sufficient for the demands of more complex texts and learning tasks of later schooling.

This panel seeks to address this gap. All studies in this panel are part of the Learning for All (L4ALL) project, a multi-site, multi-disciplinary, multi-method investigation of mid-adolescents’ literacy driven by two questions: Why does so little learning take place systematically in some schools globally? What are promising local and global levers to improve learning for all?

The L4ALL project takes place in developing countries in which advances in educational quality lag considerably behind substantive economic progress, i.e., countries with economic resources to improve educational quality but whose students still display poor educational outcomes. This project adopts the same sampling strategy across sites. Different from approaches that study exceptionally successful schools, which are often unique and hard to replicate, our approach studies “regular” public schools serving the same communities and operating under comparable circumstances. Within each developing country, we contrast more and less successful urban public schools, that are otherwise comparable. We focus on urban settings, given rapid urbanization globally (UNICEF, 2012).

This panel adopts PISA’s definition of literacy: “understanding, using, and reflecting on written texts, in order to achieve one’s goals, to develop one’s knowledge and potential, and to participate in society” (OECD, 2010, p. 37). This panel will focus not only on literacy as product, measured through standardized assessments, but also on literacy as dynamic classroom practices.

The panel format will be as follows:
- Three papers: a 20-minute presentation plus 3 minutes for clarification questions per paper; plus 2 minutes for transition time between papers (75 minutes).
- Discussant and Q&A: a 5-minute commentary by the discussant plus 10 minutes for questions (15 minutes).

Brief overview of each paper:

1. In Search of Learning for All in Botswana: Equality of Opportunities, Inequity of Outcomes. Informed by Botswana mid-adolescents’ literacy assessment results, this study examined school- and classroom-level roadblocks that interfered with learning by examining teachers’ and students’ interviews, surveys and classroom-observation data. Whereas most participating students were fluent decoders in English (the language of instruction), teachers were not equipped to address the within-classroom heterogeneity in meaning-making skills (i.e., understanding word meanings). Despite teachers and students overwhelmingly reporting mutual relationships based on trust and care, the “one size fits all” approach to learning emerged as inadequate to support individual differences.

2. Mastering the language of school: A new component to understand literacy underperformance in Brazil and Colombia. To better understand why so many adolescents struggle with literacy, this study administered a comprehensive battery of literacy assessments to Brazilian Portuguese-speaking (n=1,269) and Colombian Spanish-speaking mid-adolescents (n=1,410), similarly distributed across grades 4, 6 and 8 in poor-resourced urban public schools in São Paulo and Medellín, respectively. Results were strikingly consistent across the two sites. Beyond the effects of socio-demographic factors and word recognition skills, language proficiency (as measured by innovative previously validated Portuguese and Spanish assessments) emerged as a significant factor in explaining mid-adolescents’ text comprehension difficulties.

3. Differentiated instruction and student learning: Cross-country descriptive evidence.
Many education policies and interventions in low- and middle-income countries are motivated by a belief that instruction is not appropriately differentiated for students with vastly different prior learning levels. Using a score conversion developed by Patel and Sandefur (2020), this study examines heterogeneity in student reading outcomes within classrooms on a valid and comparable scale across 74 high, middle, and low-income countries. Then, using overlap in the administration of PIRLS items, we estimate a linking function to place students from L4All countries on the same scale as our cross-country analysis. This provides context for a rich, descriptive analysis of the instructional experiences of 4th, 6th, and 8th graders in the 3 L4All countries. We find large differences in the extent to which teachers report using differentiated instructional practices across grades and countries. This study aims to inform the wider policy context as countries consider adopting technology or curricular interventions that promote instructional differentiation.

Sub Unit

• General Pool

Individual Presentations

• Mastering the language of school: A new component to understand literacy underperformance in Brazil and Colombia. - Paola Uccelli, Harvard University

• Differentiated instruction and student learning: Cross-country descriptive evidence - Charles Gale, Harvard University

Discussant

• Karen Mundy, Toronto University