Socioeconomic achievement gaps in science and social studies in Colombia: an assessment of factors affecting school achievement

This paper investigates the factors associated with socioeconomic achievement gaps in science and social studies among high school students in Colombia and explores the implications for Education for Sustainable Development (ESD). Utilizing a comprehensive cross-cohort dataset from the SABER 11 standardized examination for the period 2017–2023, which included sample sizes ranging from 330,152–377,354 students, a four-level multilevel modeling approach was employed. Data were aggregated at the student, school, municipality, and department levels. The analysis reveals that school characteristics are the most significant factors contributing to the achievement gap, with the length of the school day and school location emerging as key predictors of academic performance in these subjects. Furthermore, the study confirms the persistence of gender and ethnic-based academic disparities. Individual socioeconomic factors, specifically insufficient nutrition, student employment, and a lack of reading habits, were found to be negatively associated with academic achievement. By shedding light on these disparities in subjects critical for ESD in the Colombian context, this research provides valuable evidence for informing and strengthening educational policies aimed at fostering more equitable learning outcomes and advancing sustainable development practices.

1 Introduction

Sustainable development (SD) aims at promoting social and economic welfare in a manner that is respectful to the environment and human culture (Sachs et al., 2024). To achieve this, it is crucial to prepare future generations with the skills and knowledge to become active members of their communities and contribute to solving some of the most pressing global issues of our time, such as climate change, poverty, and inequality. Education for Sustainable Development (ESD) constitutes a key pathway to attain sustainable development by empowering children and youth to become agents of change (Akinsemolu and Onyeaka, 2025; Kioupi and Voulvoulis, 2019; Olsson et al., 2021; Wals and Kief, 2010). Using innovative pedagogical approaches such as systems thinking, project-based learning (PBL), experiential learning, and design thinking, ESD promises to transform education from a results-based approach, where learning is assessed primarily on performance, to a learning ecosystem that envisions learning as a lifelong process essential for navigating complex global challenges (Akinsemolu and Onyeaka, 2025; O’Flaherty and Liddy, 2018; Reimers, 2024).

The goal of ESD is, therefore, to help students unlock their potential by providing them with cognitive and socioemotional tools not only to understand global issues, but also to develop their agency as they become aware of their pivotal role in society (Akinsemolu and Onyeaka, 2025; Olsson et al., 2021; Wals and Kief, 2010; UNESCO, 2020). This entails that what is learned in the classroom must be tailored to set the foundational knowledge and skills needed not only to be productive members of society, but also to actively participate in social and political life. Achieving this, however, is a complex endeavor involving systemic educational transformations. This is why UNESCO defined five priority action areas to assist countries in the process of adjusting their education systems to become more sustainable-oriented (UNESCO, 2018, 2020). From the evaluation of best practices in education, strengthening learning environments, and building the capacities of educators, to community-and-youth-based programs, these priority areas provide a roadmap for the implementation of initiatives that foster education for sustainable development.

Colombia has made important strides to transform its education system in alignment with these global objectives. For example, the Ministry of Education has formulated a series of policy guidelines for curriculum design that explicitly include topics such as restorative practices, development studies, biodiversity conservation, and human rights (Ministry of Education of Colombia, 2006, 2016a, 2016b). To assess whether students are attaining learning outcomes, the Instituto Colombiano para la Evaluación de la Educación (ICFES), the agency responsible for measuring educational quality, developed a standardized test called SABER 11. This test is administered at the end of the 11^th grade to all students in Colombia and assesses their knowledge and skills across five subjects: mathematics, reading, science, social studies, and English (Instituto Colombiano para la Evaluación de la Educación – ICFES, 2023).

Of particular interest to this study are achievement gaps in science and social studies, as their curricula and assessment frameworks are designed to unequivocally align with and integrate key goals and themes of Education for Sustainable Development (ESD). In science, for example, students are evaluated on critical aspects of environmental education such as deforestation and sustainability (Instituto Colombiano para la Evaluación de la Educación – ICFES, 2019a). In social studies, students engage with topics associated with development studies, including poverty, equity, human rights, and peace for development (Díaz Sánchez, 2020; Instituto Colombiano para la Evaluación de la Educación – ICFES, 2019b). Following the 2016 peace agreement between the Colombian government and the Revolutionary Armed Forces of Colombia (Fuerzas Armadas Revolucionarias de Colombia – FARC), a Marxist guerrilla group involved in the armed conflict, schools are mandated to teach a course on peace and conflict resolution (Vásquez-Russi, 2020)¹. Therefore, the SABER 11 test for science and social studies, by assessing content deeply integrated with ESD principles, constitutes a valuable proxy for understanding student achievement in areas critical to sustainable development efforts in Colombia.

While the prevalence of socioeconomic achievement gaps in core subjects like mathematics and reading is well-documented both internationally (Chmielewski, 2019; Coleman et al., 1966; Hanushek et al., 2022; OECD, 2023; World Bank, 2020) and in Colombia (Cabra, 2023, 2024; Instituto Colombiano para la Evaluación de la Educación – ICFES, 2023; Rodríguez-Gómez et al., 2024), less attention has been given to such disparities in subjects with explicit ESD relevance. To our knowledge, this is one of the few studies that attempts to elucidate evidence on socioeconomic achievement gaps in science and social studies for Colombia specifically, examining two academic outcomes critical for fostering sustainable development competencies and using large samples of data from multiple cohorts. This study addresses two key research questions: (1) Is there a socioeconomic achievement gap in science and social studies in Colombia? and (2) What individual and school-level factors are associated with this socioeconomic achievement gap? By utilizing comprehensive data from the SABER 11 test for the years 2017–2023 and employing multilevel modeling, this study explores the factors associated with academic achievement disparities in these ESD-relevant subjects within the Colombian context. This research contributes to the existing international literature by providing novel empirical evidence on socioeconomic achievement gaps in non-traditional, but ESD-aligned, academic domains from a Latin American country. It also offers valuable insights into the multidimensional nature of educational inequality and its implications for advancing sustainable development goals. This study is organized as follows: Section 2 presents a review of the most relevant literature on the topic and highlights the research gap. Section 3 details the methodology and data. Section 4 highlights results, while Section 5 presents a thorough discussion of findings. Section 6 presents the conclusions of the study and offers policy recommendations.

2 Literature review

Research in inequality of educational opportunity has extensively documented the existence of academic achievement gaps, often emerging as early as preschool age and persisting throughout elementary and secondary education if left unattended (Attanasio et al., 2023). While this body of literature is substantial, a significant portion has historically focused on academic disparities in mathematics and reading, considered foundational subjects and key indicators for assessing students’ core learning outcomes (Chmielewski, 2019; World Bank, 2018). Overall, these studies highlight several interconnected factors that contribute to the emergence and widening of these gaps, including growing socioeconomic disparities among children (World Bank, 2018; World Bank, UNESCO, UNICEF, 2021) and disruptions such as the recent COVID-19 crisis (Di Pietro, 2023). In what follows, we provide an overview of the theoretical and empirical literature associated with socioeconomic achievement gaps and its relevance to the Colombian context.

2.1 Socioeconomic achievement gaps: a closer look at the theory of capitals

Research on learning variability indicates that academic achievement gaps can be attributed to within-person variability, within-group variability, and contextual variability (Chmielewski, 2019; Jacobs Foundation White Paper, 2023). The Coleman et al. (1966) report is the first study that attempted to explain the factors that are associated with inequality of educational opportunity. The report found that besides teacher and school quality, socioeconomic status, conceptualized as all the social and economic assets that families have access to and command over, was the single most important predictor of academic achievement Coleman et al., 1966. Examples of these types of assets include parental education, parental employment, access to healthy food and durable goods, and access to nurturing environments where children can thrive (Black et al., 2017; World Bank, 2018).

Understanding the underlying mechanisms of these disparities requires drawing on robust theoretical frameworks from welfare economics and education. Human capital theory, for instance, suggests that investing in early child education produces the highest social and economic returns because it enables people to achieve their full potential, become productive, and increase their earnings potential (Heckman and Mosso, 2014). From this perspective, socioeconomic disparities can translate into unequal access to quality educational resources and opportunities, which can lead to gaps in the accumulation of social and economic capital. By the same token, cultural capital theory, developed by Bourdieu (1986), highlights that schools (and education systems) often reward and reinforce the cultural norms, values, and linguistic styles prevalent in dominant social classes. Because students from lower socioeconomic backgrounds possess different forms of cultural capital, this may lead to a mismatch with institutional expectations, thus, contributing to widening socioeconomic academic achievement gaps. This may explain why indigenous communities, which have a different cultural and social understanding of the world, usually perform at a lower rate in academic tests than their non-indigenous peers.

Similarly, the theory of social capital, as elaborated by scholar such as James Coleman, stresses the pivotal role that social networks and relationships within families and communities play in children’s development (Coleman, 1988). According to this theoretical framework, strong social connections, such as active parental involvement and close community partnerships, can provide resources and support to enhance children’s educational experiences, whereas a lack of such assets can exacerbate inequalities. In addition to the theories of human capital and economic and social welfare, Bronfenbrenner’s (1979) bioecological systems theory constitutes a holistic framework to understand learning variability because it shows how child development and educational outcomes are influenced by the complex interplay of multiple systems, from the immediate interactions within the family and school (microsystems), to broader societal structures and policies (macrosystems). This multi-level perspective is particularly relevant for understanding how various socioeconomic factors at different levels may contribute to widen or reduce achievement gaps (Jacobs Foundation White Paper, 2023).

2.2 Empirical evidence on socioeconomic academic achievement gaps

Studies from around the globe highlight the persistence of socioeconomic academic achievement gaps, hence, corroborating the aforementioned theoretical perspectives. Large-scale international assessments, such as the Programme for International Student Assessment (PISA) conducted by the Organisation for Economic Co-operation and Development (OECD), underpin the relationship between students’ socioeconomic status (SES) and their academic performance (OECD, 2023). Data from PISA assessments indicate, for example, that students from marginalized identities perform lower academically than their more affluent peers, even across wealthy countries (OECD, 2023). Similarly, the Trends in International Mathematics and Science Study (TIMSS) also report significant socioeconomic and gender-based academic achievement gaps in mathematics and science across countries (Broer et al., 2019). Besides these large-scale assessments, national studies from various developed and developing countries further explicate how socioeconomic factors impact learning. For example, research in the United States underscores the role of school funding in intensifying academic disparities (National Academies of Sciences, Engineering, and Medicine, 2023), while studies in other Latin American countries point to the impact of factors such as rural–urban divides (Rodríguez-Gómez et al., 2024), inequitable access to information durable goods (Cabra, 2023, 2024), nutrition (Prangthip et al., 2019), and length of school day (Alfaro et al., 2015) on educational outcomes. Together, these studies underscore the complex and multifaceted nature of educational inequality.

A key factor identified across this literature is parental involvement in students’ academic and extracurricular activities. Scholars such as Castro et al. (2015) have demonstrated that students whose parents are more actively involved in their school life tend to perform better academically. Another critical factor is the impact of school quality and teachers on student performance. Researchers like Schneider (2017) suggest that poor school quality, characterized by factors such as a lack of educational resources or a violent school climate, has a negative impact on academic achievement. Similarly, studies on the effect of teachers’ leadership and engagement indicate that students learn more and perform better academically when their educators are fully committed to their learning process (World Bank, 2018; Liu, 2024).

However, despite this extensive research on general academic achievement gaps, there is limited empirical evidence specifically exploring socioeconomic disparities in academic outcomes related to Education for Sustainable Development (ESD). This gap in the literature is partially explained by the inherent difficulties associated with effectively measuring attainment in this area. ESD, as conceived by UNESCO, is understood as an approach to education that seeks to empower children and youth to contribute to addressing society’s most pressing global challenges, including poverty, climate change, racism, and gender-based discrimination (Glavič, 2020; O’Flaherty and Liddy, 2018; UNESCO, 2020). Effective assessment in ESD must consider not only the acquisition of knowledge and skills, but also the development of agency in the application of such abilities in specific contexts (UNESCO, 2018, 2020; Tikly, 2019). Consequently, much of the existing research on ESD takes the form of case studies that showcase the work of schools, community-based organizations, and youth-led projects in promoting ESD initiatives, with a big share of it taking place in the African and Asian regions (Lee and Fujii, 2024a; Lotz-Sisitka and Lupele, 2017; Sachs et al., 2019; UNESCO, 2020; Wals and Kief, 2010). While these studies are pivotal to our understanding of effective practices in ESD, they generally do not allow for a broader assessment of progress, or lack thereof, in the implementation of ESD guidelines at a systemic or policy level (Lee and Fujii, 2024b; Uleanya, 2025).

Colombia provides a compelling case study of strong institutional leadership in promoting ESD. Since 1998, with the implementation of the “Curricular Guidelines” by the Ministry of Education, the country has actively sought to integrate education with sustainable development principles. This has resulted in the design of a national framework for teaching key topics across the K-12 setting, including strategies to foster peaceful conflict resolution. Strengthened by additional policy frameworks such as the “Standards of Basic Learning Outcomes” and the “Basic Learning Rights,” the Colombian government has demonstrated its commitment to providing an education oriented towards cultivating the principles of sustainable development (Ministry of Education of Colombia, 2016a, 2016b). Despite these significant policy efforts, reports indicate that for the period 2014–2023, academic achievement in science and social studies has remained relatively low, with students performing at a below average level (Laboratorio de Economía de la Educación (LEE) de la Pontificia Universidad Javeriana, 2024). In science, for example, students’ average score for the period 2014–2023 was 49.69 (out of a total of 100 points). In contrast, the average score for social studies for the same period was 48.61. Although useful, these reports have only provided a partial picture of the underlying nature of the relationship between academic achievement in ESD-relevant subjects and socioeconomic disparities. Therefore, there remains a notable gap in large-scale empirical evidence regarding how socioeconomic factors influence student achievement in subjects explicitly aligned with ESD in Colombia. This study addresses this gap by providing novel evidence on socioeconomic achievement gaps in science and social studies for Colombia. By using comprehensive cross-cohort data from the SABER 11 test across multiple years and employing robust multilevel modeling, this research contributes to our understanding of the achievement gap in these ESD-relevant subjects within the Colombian context, and offers insights relevant to broader discussions on educational equity and sustainable development globally.

3 Methodology and data

This study analyzed the factors associated with the socioeconomic achievement gap in science and social studies in Colombia from 2017 to 2023. Multilevel modeling was used as the method to estimate this gap for two reasons. First, multilevel modeling is a type of multiple regression analysis that considers when the data are structured at various hierarchical levels (Raudenbush et al., 2019). Not accounting for this clustering factor could yield biased estimators (Raudenbush et al., 2019). Second, effect sizes of multilevel modeling are more efficient than the ones produced by multiple regression analysis because they take into account the variance at the different levels of aggregation of the data.

For this study, a four-level model was constructed because data were grouped at four levels: students (level 1), schools (level 2), municipalities (level 3), and departments (level 4). The study used the estimation model proposed by Cabra, 2023:

$\begin{array}{ll}{Y}_{\mathit{\text{ijkl}}}={\beta }_0+{\beta }_1{X}_{\mathit{\text{ijkl}}}+(e_{0\mathit{\text{ijkl}}}+u_{0\mathit{jkl}}+v_{0\mathit{kl}}+f_{0l})& (1)\end{array}$

In Equation 1, $Y_{\mathit{\text{ijkl}}}$ corresponds to one of the academic outcomes for student i grouped in school j, municipality k, and department l. In the model, $X_{\mathit{\text{ijkl}}}$is a vector of covariates, and the random effects $e_{0\mathit{\text{ijkl}}}$ , $u_{0\mathit{jkl}}$ , $v_{0\mathit{kl}}$ , and $f_{0l}$ are the residual differentials for each of the levels (students, schools, municipalities, and departments). In this four-level model, the residual differentials for students, schools, municipalities, and departments are assumed to be normally distributed (Cabra, 2023).

3.1 Data

Data for this study were provided by the Instituto Colombiano para la Evaluación de la Educación (ICFES), a government agency responsible for assessing student performance across different subjects. The data collected corresponded to the SABER 11 test, a nationwide exam administered to students in their last year of high school, for the period 2017–2023. Although all high school students must take the SABER 11 test, only students with complete information were included in the analytical sample. Because less than 5% of data were missing completely at random, we employed listwise deletion as the method for dealing with missing values. As such, we have an analytical sample that is smaller than the population of high school students that took the exam for each of the cohorts. We selected this period of study because 2017 was the first year where students were assessed on topics associated with peace, development studies and conflict resolution as part of the social studies test. Similarly, we selected 2023 as the last cohort because that was the last year for which information was fully available. Hence, the inclusion of data from 2017 to 2023 allows for a comprehensive cross-cohort analysis, enabling us to observe trends and the persistence of socioeconomic achievement gaps over time, including the period affected by the COVID-19 pandemic. While the stability of mean scores across these years might appear surprising given the pandemic’s impact on education, this study explores the factors associated with achievement gaps in science and social studies within each cohort. Table 1 summarizes the basic characteristics of the datasets used for this study, including scores for school outcomes and sample sizes.

Table 1

Table 1. Data, school outcomes, and sample sizes.

As observed, mean scores for science and social studies have remained similar from year to year, with students performing slightly higher in science. Similarly, sample sizes have ranged from 330,152 students in 2023 to 377,354 in 2020. The large samples reported in Table 1 can contribute to reducing the margin of error of the estimation methods, hence, providing more robust and accurate findings.

3.2 Participants

In Colombia, school size varies by region and school type, with most students attending public schools located in urban areas (more than 75% of the student population). According to a report by the OECD (2025), the median primary school in Colombia serves about 30 students per grade. The samples used for this study consisted of seven cohorts (years 2017–2023). Students from each of the cohorts are in their last year of high school, which is the year where they must take the SABER 11 test. The number of participants ranged from 330,152 students for the year 2023 to 377,354 for the year 2020. The age of the participants ranged from 15 to 17 years old across samples. To minimize any potential risks to participants, data were de-identified.

3.3 Variables

The variables used in this study captured information on academic achievement in science and social studies for students in 11^th grade in Colombia, sociodemographic and socioeconomic characteristics of students, and information about schools. Below, we provide a detailed description of the different types of variables used in this study.

3.3.1 Dependent variables

This study used two dependent variables to measure academic achievement in subjects relevant to Education for Sustainable Development: (1) results from the SABER 11 test for science and (2) results from the SABER 11 test for social studies. These variables were selected because their curricula and assessment content in Colombia explicitly align with and integrate key themes and competencies emphasized by the guidelines of UNESCO regarding the scope of ESD. Additionally, and given that much of the research agenda on socioeconomic academic achievement gaps in Colombia has primarily focused on reading and math literacy, this study sought to amplify its scope to other content areas with direct implications for education for sustainable development. Test scores were standardized to facilitate interpretation of results.

3.3.2 Covariates

Covariates were selected based on research on equity in education and socioeconomic achievement gaps, which points to gender, ethnic identity, parental education, socioeconomic status, access to durable goods, and school characteristics as highly associated with school performance. In what follows, we explain each of the covariates used in this study.

Gender. Research on academic achievement indicates disparities by gender, with boys performing better than girls in Science, Technology, Engineering, and Mathematics (STEM)-related subjects (Meinck and Brese, 2019). This variable was included to assess its association with academic achievement in ESD-relevant subjects. We coded this variable as “0” for boys and “1” for girls.

Ethnic identity. Studies on racial equity in education suggest the existence of ethnic and racial academic disparities, disproportionately affecting students of color. Studies show, for example, that in the United States, students who identify as Latinos or African Americans perform at a lower rate than their White peers (Morgan et al., 2023). This variable was included to explore its association with academic achievement in ESD-relevant subjects. We coded this variable as “0” for students who did not identify as belonging to an ethnic group and “1” otherwise.

Mother’s education. Research on child development indicates that mother’s education has a positive impact on various development outcomes for children. For example, empirical evidence shows that higher levels of maternal education are highly associated with higher school enrollment, better academic performance, and less truancy (Wang et al., 2020). We used this variable to assess whether higher levels of maternal education were associated with better academic performance in ESD. We coded this variable into six categories: 0 = no education, 1 = elementary education, 2 = secondary education, 3 = technical and vocational education, 4 = college education, 5 = graduate education.

Socioeconomic status (SES). Research on socioeconomic achievement gaps indicates that students who experience high poverty perform lower than students from affluent backgrounds (Coleman et al., 1966; von Stumm et al., 2022). We used this variable to understand if academic disparities in ESD varied by socioeconomic status. Data used for this study captured a variable that measures socioeconomic status for Colombian children. This is a wealth index that comprises possession of durable goods, parental education, and parental employment. Students are then distributed in quartiles depending on their score on the wealth index. As such, this variable was coded as: 1 = low SES, 2 = lower-middle SES, 3 = middle SES, 4 = high SES.

Dietary intake. Studies on equity of opportunity in education highlight that nutrition is a key indicator of academic performance (Prangthip et al., 2019). We used this variable to assess whether high levels of nutrition, as measured by consumption of vegetables, dairy products, and protein on a weekly basis, would influence academic achievement in ESD. We coded this variable as: 0 = never or seldom, 1 = once or twice a week, 2 = more than three times a week.

Information goods. Research on the impact of durable goods in educational outcomes indicates that computers and access to the internet have a positive effect on academic achievement in reading and math (Cabra, 2023, 2024). We used this variable to assess whether the possession of information goods (computer and access to internet at home) was related to academic achievement in ESD. This variable was coded as “0” if students did not have these commodities and “1” if they did have them.

Reading for fun. Studies on academic achievement suggest that students who engage in activities that stimulate their learning, such as reading for pleasure in their spare time do better academically in school than students who do not (Sun et al., 2024). We coded this variable as: 0 = does not read for fun, 1 = spends 30 min or less reading for fun, 2 = spends less than 2 h reading for fun, 3 = spends more than 3 h reading for fun.

Time spent on the internet. Research on the impact of Information and Communications Technology (ICT) in school performance indicates that targeted use of ICT resources is highly related to improved academic achievement (Ladrón de Guevara et al., 2022). We coded this variable as: 0 = does not use the internet, 1 = spends 30 min or less on the internet, 2 = spends less than 3 h on the internet, 4 = spends more than 3 h on the internet.

Child labor. Research on poverty and equity in education demonstrates that child labor is negatively associated with school outcomes (Vasey, 2020). Although there are many forms of child labor, in this study this variable was used to measure whether students held a paid job while attending school. We coded this variable as “0” if students did not have a paid job and “1” otherwise.

School type. Research on school quality and academic performance indicates that students who attend private schools perform better than students who attend public schools (OECD, 2024). This variable was coded as “0” for private schools and “1” public schools.

School setting. Studies on equity of opportunity in education stress the existence of academic achievement gaps by school location. In general, research shows that children who attend schools in urban areas perform better academically compared to children in rural schools (Bonilla-Mejía and Londoño-Ortega, 2021). This variable was coded as “0” for urban schools and “1” for rural schools.

Length of school day. Research on instruction time and school outcomes suggests that longer school days have a positive effect on academic achievement, school attendance, and social outcomes (Alfaro et al., 2015; Hincapie, 2016). This variable is relevant for this study because Colombia has made a commitment to transition from a half-school-day format (less than 5 h of classes a day) to a full-school-day format (more than 7 h of class a day). This variable was coded: 1 = continuing education (students who dropped out of school and are completing their GED), 2 = half day format, and 3 = full day format.

4 Results

Results are shown in the order that we conducted data analysis. Therefore, we present the descriptive statistics first, followed by the correlation and the multivariate regression analyses. Finally, we present the results of the multilevel linear estimation.

4.1 Descriptive statistics

Table 2 presents the sociodemographic information of students in the samples as well as the characteristics of the schools. Across samples, about 54% of students identified as girls and less than 7% identified as belonging to an ethnic group. Regarding maternal education, most of the students in the samples had mothers who only completed elementary education (about 25%) and secondary education (roughly 42%), and less than 15% of the students across the samples had mothers who finished college. In terms of SES, most students across the samples identified as lower-middle income (close to 38%) and middle income (close to 40%). In relation to nutrition levels, most students across the samples reported that they never or seldom eat vegetables, dairy products, and animal protein during the week (about 60%) compared to students who have access to those foods more than three times a week (30%). Regarding information goods, the percentage of students who had access to computers and the internet has increased over the years, particularly during the pandemic (reaching 55% across the samples). Additionally, 80% of students across the samples reported reading for fun for less than 2 h a day while more than 50% indicated using the internet for less than 3 h a day. In relation to paid child labor, close to 36% of the students across the samples reported having a paid job, a percentage that increased since 2020.

Table 2

Table 2. Sociodemographic characteristics of students and information about schools.

Regarding school characteristics, close to 76% of the students across the samples attended public schools and more than 80% attended schools in urban areas. In relation to length of school day, data for the different cohorts indicated an increase in the percentage of students who attended full school day programs, going from 26% in 2017 to roughly 45% in 2023. The descriptive information reported in Table 2 indicates well-balanced samples for each of the cohorts used in the study as well as for the variables included in the estimation models. This is crucial for conducting advanced statistical analysis because it guarantees there is enough statistical power to detect significant relationships between the variables of interest. The observed increase in full day school attendance in 2020, despite the pandemic, obeys to a government policy that requires schools to gradually transition from half day to full day programs (Consejo Nacional de Política Económica y Social – CONPES, 2015). Also, it is important to stress that school closures started towards the end of march, so that number reflects the net enrollment before the pandemic. Once schools were shut down, the number of students registered as attending a full day program did not change in the government records.

4.2 Correlation and multivariate regression analyses

A Pearson correlation analysis was used to measure the level of interdependence between the variables of interest. Most variables in the study had a correlation coefficient below 0.3, indicating a weak correlation. Socioeconomic status (SES) and possession of information goods had a correlation coefficient of 0.7, suggesting a strong association (see Appendices 1–7)².

We conducted multivariate regression analyses for each cohort to test whether the selected variables had a statistically significant relationship with the outcomes of interest. Across samples, adjusted $R^2$ accounted roughly for 60% of the variance in science and social studies scores, which indicated that the variables included in the models were good predictors of the test results in science and social studies (see Appendices 8–14) (see footnote 2).

4.3 Multilevel analysis

We performed multilevel analysis to understand the factors that are related to academic achievement in science and social studies in Colombia. In multilevel analysis, we start by building a null model to estimate whether the data show evidence of clustering or group effects. Across samples, results from the null models for each cohort indicated the existence of group effects, with most of the variance in science and social studies explained at the school and department levels (see Appendices 15, 16) (see footnote 2). The Intraclass Correlation Coefficient (ICC) for the null models, an indicator of how much of the total variance in the dependent variables is attributable to the grouping effect, suggested that school, municipality, and department random effects constituted roughly 30% of the total residual variance across samples (see Appendices 15, 16).

In what follows, we present the results from the multilevel models for science and social studies. Tables 3, 4 summarize the results of the estimation models. Table 3 presents the results for science and Table 4 the results for social studies. It is important to stress that the data used for this study is not longitudinal but cross-sectional, and thus, causation cannot be inferred. However, the associations found among the variables of interest have important research and policy implications regarding the multidimensional nature of socioeconomic academic disparities, particularly when analyzing content areas different from reading and mathematics.

Table 3

Table 3. Multilevel linear models for science.

Table 4

Table 4. Multilevel linear models for social studies.

4.3.1 Results for science

Results indicated the existence of academic disparities associated with gender and ethnic identity. Across samples, girls performed lower than boys in science, with an average effect size of 0.23. Similarly, students who identified as belonging to an ethnic group scored, on average, lower in science than students who did not identify as members of an ethnic group (average effect size of 0.11). Additionally, students who worked while attending school scored, on average, lower in science than students who did not work (average effect size of 0.17).

Furthermore, results suggested that maternal education, nutrition, access to information goods, reading for pleasure, time spent on internet, and SES were positively associated with science scores (an ESD-relevant subject). For example, coefficients were larger for students whose mothers had attained higher levels of education (average effect size of 0.09) and for students who reported a rich diet (average effect size of 0.04). Similarly, coefficients were larger for students who had access to information goods, for students who read for fun more often, and for students who spent more time on the internet (average effect sizes of 0.05, 0.14, and 0.05, respectively). SES was positively associated with science scores for the period 2017–2020 and for the year 2023, with an average effect size of 0.03, but not statistically significant for the years 2021 and 2022.

Regarding school characteristics, results showed that children who attended public schools and schools in rural areas performed lower in science than students who attended private schools and schools in urban settings (average effect sizes of 0.19 and 0.17, respectively). In contrast, the models found that length of school day was positively associated with science scores across all samples. As such, students who attended full day programs scored, on average, higher than students who enrolled in half-day or continuing education programs (average effect size of 0.34).

4.3.2 Results for social studies

The models for social studies suggested the persistence of equity achievement gaps. For example, results indicated that girls scored, on average, lower than boys in social studies (average effect size of 0.14). Similarly, students who identified as members of an ethnic group scored lower than students who did not report belonging to an ethnic group (average effect of 0.12). Moreover, students who held a paid job while completing their studies performed lower in social studies than students who did not work, with an average effect size of 0.19.

In relation to socioeconomic factors, findings indicated that maternal education, nutrition, access to information goods, reading for fun, using the internet, and SES had a positive association with achievement in social studies. Regarding mother’s education and nutrition, for instance, coefficients were larger for students whose mothers had more education and for students with a healthy and regular diet (average effect sizes of 0.09 and 0.04, respectively). Also, coefficients were larger for children who had access to computers and internet, for students who reported reading regularly, and for students who used the internet more often (average effect size 0.04, 0.17, and 0.05, respectively). SES was positively associated with social studies scores for the period 2017–2019, with an average effect size of 0.02, but negatively related for the period 2021–2023, with an average effect size of 0.05.

As to school characteristics, results indicated that performance in social studies was lower for public schools (average effect size of 0.21) and for rural schools (average effect size of 0.18) in comparison to private schools and schools in urban settings. Conversely, the models found a positive association between length of school day and social studies scores across all samples. Thus, students who attended full day programs scored, on average, higher in the social studies test compared to students who enrolled in half-day or continuing education programs (average effect size of 0.29).

5 Discussion

This study provides robust empirical evidence on the factors associated with socioeconomic academic achievement gaps in science and social studies among 11^th grade students in Colombia for the period 2017–2023. By using multilevel modeling and academic outcomes different from language and mathematics, this research contributes to the existing literature on socioeconomic achievement gaps by showing that significant disparities persist across different individual and school-level characteristics in academic domains explicitly aligned with ESD principles and themes.

Regarding individual-level factors, our findings confirm the persistence of gender and ethnic-based academic disparities in Colombia. These results are consistent with international research on gender gaps in STEM education (Meinck and Brese, 2019) and racial academic disparities in the United States (Morgan et al., 2023). The lower performance of girls in science, and ethnic minority students in both science and social studies, is particularly concerning within the framework of ESD. While Colombian curricula for these subjects emphasize equity and indigenous knowledge, our results suggest that factors outside the formal curriculum, such as family expectations and the differential distribution of cultural capital and social capital (Bourdieu, 1986; Coleman et al., 1966; Coleman, 1988), may impede the full realization of these students’ potential to engage with and contribute to solving sustainable development issues. These findings underscore the need for targeted interventions that address systemic barriers and foster inclusive learning environments, drawing lessons from international efforts to mitigate similar disparities.

Furthermore, the negative association between child labor and academic performance in both science and social studies aligns with extensive international research on the detrimental effects of child labor on educational attainment (Vasey, 2020; World Bank, 2018). This finding highlights the severe impact of poverty on child development and education because it forces children to assume financial responsibilities that compromise their learning time and school engagement. From a human capital perspective, participation in child labor reduces the investment in education, which may lead to low skill development. This issue is particularly severe in high poverty contexts, and our findings for Colombia resonate with similar challenges observed across developing countries. Evidence from this study emphasized the urgency to formulate social protection policies that allow children to focus on their education, a fundamental prerequisite for their future contributions to sustainable development.

Our results also corroborate previous research on the positive association among maternal education, household wealth (SES and possession of durable goods), nutrition, reading habits and academic achievement (Cabra, 2023, 2024; Prangthip et al., 2019; von Stumm et al., 2022; Wang et al., 2020; World Bank, 2018; Sun et al., 2024). In particular, maternal education and SES align with human capital and cultural capital theories, suggesting that children whose parents have higher levels of education and more access to assets have more opportunities to engage in learning. Similarly, the strong positive association between adequate nutrition, access to information goods (computers and internet) and academic achievement also underscores the importance of fulfilling basic needs as a foundational resource to enhance learning. This is particularly relevant in the context of global efforts to reduce educational inequalities, where promoting healthy development and access to technology are considered crucial for student success. Reading for pleasure, a socioemotional and cultural factor, not only enhances literacy skills but also cultivates cognitive and metacognitive abilities essential for deeper understanding and critical thinking, foundational skills in ESD.

In relation to school-related factors, this study identifies the length of the school day as the most important factor associated with academic achievement in science and social studies in Colombia for the period 2017–2023. This result is consistent across the cohorts studied, even during the COVID-19 pandemic. A potential explanation for the observed increase in the number of students attending full day programs in 2020 corresponds to the government’s efforts to implement the full day school policy (Consejo Nacional de Política Económica y Social – CONPES, 2015). As such, the rise in the number of students attending full day programs between 2019 and 2020 may be explained by the fact that in that year the number of available seats for full day programs increased. Moreover, it is important to highlight that schools closed towards the end of march, which entails that student enrollment had been completed before the pandemic. This suggests that the increase in the number of students attending full day programs has more to do with the implementation cycle of the full day school policy (e.g., achieving outcomes, monitoring progress, or resource management) than with any irregularity in the transition to a remote education model. This finding is consistent with Alfaro et al. (2015) and Hincapie (2016), who argue that schools operating on a full day format can offer more comprehensive educational experiences to children, hence, reducing achievement gaps. A potential explanation for this finding is that longer school days often correlate with increased instructional time, more opportunities for extracurricular activities that foster socioemotional skills, and access to essential services like universal meal programs, all of which contribute to enhancing human capital development. The observed gradual decrease in the achievement gap over the study period, paired with the increase in the percentage of students who attend full day programs, shows that extending instructional time can positively impact educational equity and contribute to the implementation of ESD-relevant learning outcomes.

Furthermore, our results indicate that students in public schools and rural areas perform lower than those in private and urban schools, corroborating previous research regarding educational inequalities by school type and location (Bonilla-Mejía and Londoño-Ortega, 2021; OECD, 2024; World Bank, 2018). These disparities often stem from unequal resource distribution, including disparities in qualified teachers, infrastructure, and technological access, as noted in the Colombian context. These findings resonate with the bioecological systems theory of Bronfenbrenner (1979), which highlights that child development and learning are permeated by different interlinked factors. Addressing these gaps requires systemic interventions aimed at enhancing resource allocation and improving educational quality in underserved public and rural schools.

5.1 Implications for education for sustainable development

Our findings on the persistence of socioeconomic and equity-based achievement gaps in science and social studies carry significant implications for the advancement of ESD in Colombia and potentially for other similar contexts. While the SABER 11 test is not a direct assessment of ESD competencies, its strong alignment with ESD-related curricular content (e.g., environmental education in science, human rights, development, and peace in social studies) means that disparities in these subjects directly undermine the overarching goals of sustainable development. If substantial portions of the student population, particularly those from disadvantaged socioeconomic or ethnic backgrounds, are underperforming in these crucial areas, they are less likely to develop the critical knowledge, problem-solving skills, and agency necessary to become informed and active citizens who can participate in addressing sustainability issues. Our results emphasize that achieving equitable and high-quality education, which addresses fundamental socioeconomic barriers, is a prerequisite for effective ESD implementation. The identified factors associated with the gaps, from child labor to length of school day, constitute key indicators for policy intervention that can simultaneously promote educational equity and strengthen educational policies to advance in the implementation of sustainable development goals.

5.2 Limitations and future research

One of the main limitations of this study is that it relies on cross-sectional data, which facilitates the identification of associations but not direct causal relationships between the selected variables and academic achievement. Future studies on this area could benefit from longitudinal data that monitor student progress over time and from quasi-experimental designs to measure the impact of specific interventions aimed at reducing these academic gaps. Similarly, if future research aims to establish causality in the absence of longitudinal or experimental data, the application of instrumental variables could be a valuable methodological consideration. Another limitation of the study is the lack of a decomposition analysis to assess the effects of gender and ethnicity. Future research could employ decomposition models such as the Oaxaca-Blinder decomposition to further explore whether the observed effects of gender and ethnic identity are solely due to these factors or to other elements associated with them. This approach could provide a more nuanced understanding of the mechanisms driving these disparities, even with the low correlations observed among the explanatory factors. Moreover, while we demonstrate a strong alignment between the assessed subjects and ESD principles, the development and validation of robust tools to measure comprehensive ESD competencies would significantly advance the field.

6 Conclusion

This study provides robust empirical evidence on the existence of socioeconomic and equity-based academic achievement gaps in science and social studies among Colombian high school students and their implications for Education for Sustainable Development (ESD). Our findings, derived from multilevel modeling analysis of the SABER 11 test for high school students for the period 2017–2023, reveal that school characteristics are significant factors associated with these disparities. In particular, the length of the school day emerged as the most salient factor associated with the academic achievement gap in these ESD-relevant subjects. Similarly, the study showed academic disparities between rural versus urban schools and public versus private schools, underscoring the pressing need to address school-level inequities at the policy level. Moreover, this research showed the persistence of gender and ethnic-based academic achievement gaps in science and social studies, suggesting deep-rooted structural equity issues that require holistic social policy interventions.

Furthermore, our findings offer valuable insights for policymakers in Colombia and other contexts facing similar educational challenges. Evidence from this study suggests that designing education systems that acknowledge and actively address learning variability in student and school contexts is crucial for addressing inequalities. This may entail strategically extending the length of the school day for all schools, particularly in underserved areas, increasing funding and resources for rural and public schools, and providing out-of-school learning opportunities designed to mitigate gender and ethnic-based gaps.

This study carries important methodological and policy implications for the broader field of education for sustainable development. Methodologically, it highlights the potential for utilizing existing standardized assessments, whose curricula align with ESD principles, to monitor progress on ESD-relevant learning outcomes in the absence of specific ESD assessment tools. It also implies that future cross-cohort studies, particularly those in contexts similar to Colombia, should aim at the development of more comprehensive evaluation tools to assess full achievement in ESD, particularly regarding the assessment of attitudes and competencies that promote sustainable development. Similarly, future research on academic achievement in ESD-relevant subjects would benefit from including a qualitative component to further explicate the complex interplay of variables such as gender and ethnic identity on performance.

At the policy level, results from this study stress that a fundamental step for governments to align their education systems with the goals of ESD is to systematically review and ensure consistency between their national curricula and their assessment systems. This iterative process will allow them to understand whether content taught in schools, especially in subjects like science and social studies, is effectively measured at the end of the education cycle, and whether it cultivates the knowledge and skills necessary to promote sustainable development. Furthermore, given the current lack of a universally standardized tool to measure comprehensive ESD achievement, it is imperative to conduct cross-country comparisons on the different evaluation systems used to assess performance in subjects closely related to ESD. Findings from this study, for example, indicate that the contents and structure of the SABER 11 test for science and social studies serve as a valuable indicator of progress towards ESD-related learning. This approach could significantly help policymakers and government agencies responsible for education policy to strengthen their curricula and current evaluation systems, ensuring they are more closely aligned with the holistic learning outcomes envisioned for ESD. A starting point could be the establishment of regional and international dialogues to share best practices on curriculum design and assessment, contributing to the creation of a global ecosystem for the effective implementation and monitoring of ESD.

6.1 Policy recommendations

Based on our findings, we recommend four policy interventions to address the identified socioeconomic achievement gaps in Colombia and to strengthen ESD efforts. These include: (1) Strengthening social protection programs to reduce child labor and alleviate poverty, thus enabling children to prioritize education, (2) Expanding and equitably funding full day school programs, particularly in public and rural areas, to ensure that all students benefit from extended instructional time and more educational resources, (3) Investing strategically in rural and public schools to address resource disparities, enhance teacher training, and improve infrastructure and technological access, and (4) Developing targeted educational support programs that address the specific needs of ethnic groups and promote gender equity in STEM-related subjects. These policy interventions will contribute to creating more equitable educational opportunities, reduce academic achievement in critical subjects associated with EDS, and strengthen Colombia’s capacity to prepare the next generation of young leaders to tackle the complex challenges of sustainable development.

Data availability statement

Publicly available datasets were analyzed in this study. This data can be found at: https://icfesgovco.sharepoint.com/sites/DataIcfes2.0/SitePages/Saber11_Bases-y-documentación%20técnica.aspx.

Author contributions

HC: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing.

Funding

The author(s) declare that no financial support was received for the research and/or publication of this article.

Conflict of interest

The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author declares that no Gen AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/feduc.2025.1658707/full#supplementary-material

Footnotes

1. ^This course is called Cátedra para la Paz. In this course students learn about the history of the armed conflict in Colombia and techniques for the peaceful resolution of conflicts.

2. ^All appendices are available at https://drive.google.com/drive/folders/1WjKUEKbsGpL16qV865UiEI28Tb9MpfGU?usp=share_link

References

Akinsemolu, A., and Onyeaka, H. (2025). The role of green education in achieving the sustainable development goals: a review. Renew. Sust. Energ. Rev. 210:115239. doi: 10.1016/j.rser.2024.115239

Crossref Full Text | Google Scholar

Alfaro, P., Evans, D. K., and Holland, P. (2015). Extending the school day in Latin America and the Caribbean. World Bank Group. Education global practice group. Policy research working paper 7309. Available online at: http://documents.worldbank.org/curated/en/477421467986293530/pdf/WPS7309.pdf

Google Scholar

Attanasio, O., Lopez, F., Perez-Lopez, D., and Reynolds, S. (2023). Inequality in the early years in LAC: a comparative study of size, persistence, and policies. Interamerican Development Bank. doi: 10.18235/0005359

Crossref Full Text | Google Scholar

Black, M. M., Walker, S. P., Fernald, L. C. H., Andersen, C. T., DiGirolamo, A. M., Lu, C., et al. (2017). Early childhood development coming of age: science through the life course. Lancet 389, 77–90. doi: 10.1016/S0140-6736(16)31389-7

Crossref Full Text | Google Scholar

Bonilla-Mejía, L., and Londoño-Ortega, E. (2021). Geographic isolation and learning in rural schools. Borradores de Economía. Banco de La República de Colombia. Available online at: https://repositorio.banrep.gov.co/server/api/core/bitstreams/0a9e6cc6-89b0-4d3b-a255-03db6556b42f/content

Google Scholar

Bourdieu, P. (1986). The forms of capital, in Handbook of theory and research for the sociology of education. Greenwood Press. Available online at: https://jan.ucc.nau.edu/~sj6/Bourdieu,%20The%20Forms%20of%20Capital.pdf

Google Scholar

Broer, M., Bai, Y., and Fonseca, F. (2019). Socioeconomic inequality and educational outcomes: Evidence from twenty years of TIMSS. Cham: Springer.

Google Scholar

Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and design. Cambridge, MA: Harvard University Press.

Google Scholar

Cabra, H. (2023). Three approaches to modeling the relationship among durable goods, academic achievement, and school attendance in Colombia1. Heliyon 9:e22732. doi: 10.1016/j.heliyon.2023.e22732

Crossref Full Text | Google Scholar

Cabra, H. (2024). The role of possession of durable goods in gender-based developmentally-situated educational disparities in Colombia. Int. J. Educ. Res. Open 7:100356. doi: 10.1016/j.ijedro.2024.100356

Crossref Full Text | Google Scholar

Castro, M., Expósito-Casas, E., López-Martín, E., Lizasoain, L., Navarro-Asencio, E., and Gaviria, J. (2015). Parental involvement on student academic achievement: a meta-analysis. Educ. J. Rev. 14, 33–46. doi: 10.1016/j.edurev.2015.01.002

Crossref Full Text | Google Scholar

Chmielewski, A. K. (2019). The global increase in the socioeconomic achievement gap, 1964 to 2015. Am. Sociol. Rev. 84, 517–544. doi: 10.1177/0003122419847165

Crossref Full Text | Google Scholar

Coleman, J. (1988). Social capital in the creation of human capital. Am. J. Sociol., 94: S95–S120. Available online at: http://www.jstor.org/stable/2780243

Google Scholar

Coleman, J., Campbell, E., Hobson, C., McPartland, J., Mood, A., Weinfeld, F., et al. (1966). The equality of educational opportunity. National Center for educational statistics. Available online at: https://files.eric.ed.gov/fulltext/ED012275.pdf

Google Scholar

Consejo Nacional de Política Económica y Social – CONPES (2015). Documento Conpes 3831: Declaración de importancia estratégica del plan Nacional de Infraestructura Educativa Para la implementación de la jornada única escolar. [Conpes document no. 3831: National educational infrastructure plan for the implementation of the full day school policy]. Departamento Nacional de Planeación. Available online at: https://colaboracion.dnp.gov.co/CDT/Conpes/Económicos/3831.pdf

Google Scholar

Di Pietro, G. (2023). The impact of Covid-19 on student achievement: evidence form a recent meta-analysis. Educ. Res. Rev. 39:100530. doi: 10.1016/j.edurev.2023.100530

PubMed Abstract | Crossref Full Text | Google Scholar

Díaz Sánchez, E. (2020). Enseñanza de los derechos básicos de aprendizaje Para las ciencias sociales [teaching of the basic learning rights for social studies]. Voces y Silencios: Revista Latinoamericana de Educación 11, 46–65. doi: 10.18175/VyS11.1.2020.3

Crossref Full Text | Google Scholar

Glavič, P. (2020). Identifying key issues of education for sustainable development. Sustainability 12:6500. doi: 10.3390/su12166500

Crossref Full Text | Google Scholar

Hanushek, E., Light, J., Peterson, P., Talpey, L., and Woessmann, L. (2022). Long-run trends in the U.S. SES achievement gap. Educ. Finance Policy 17, 608–640. doi: 10.1162/edfp_a_00383

Crossref Full Text | Google Scholar

Heckman, J., and Mosso, S. (2014). The economics of human development and social mobility. Annu. Rev. Econ. 6, 689–733. doi: 10.1146/annurev-economics-080213-040753

PubMed Abstract | Crossref Full Text | Google Scholar

Hincapie, D. (2016). Do longer school days improve student achievement? Evidence from Colombia. Inter-American Development Bank. Working paper series, no. IDB-WP-679. doi: 10.18235/0000268

Crossref Full Text | Google Scholar

Instituto Colombiano para la Evaluación de la Educación – ICFES (2019a). Marco de referencia de la prueba de ciencias naturales Saber 11 [Reference framework for the SABER 11 test for science]. Available online at: https://www.icfes.gov.co/wp-content/uploads/2024/09/marco-de-referencia-prueba-de-ciencias-naturales-saber-11-1-1.pdf

Google Scholar

Instituto Colombiano para la Evaluación de la Educación – ICFES (2019b). Marco de referencia de la prueba de sociales y ciudadanas saber 11 [reference framework for the SABER 11 test for social studies and citizen competencies]. Available online at: https://www.icfes.gov.co/wp-content/uploads/2024/09/mr-sociales-y-ciudadanas-saber-11-2022.pdf

Google Scholar

Instituto Colombiano para la Evaluación de la Educación – ICFES (2023). Inequidades en el logro academico en las regiones: Resultados en la prueba de lectura [Socioeconomic achievement gaps across regions: Results from the reading test]. Available online at: https://www.icfes.gov.co/apuntes-del-icfes-para-la-politica-educativa/

Google Scholar

Jacobs Foundation White Paper (2023). “Advancing Research on Learning Variability,” in Research Agenda, eds. J. M. Tsang, G. Fetz-Fernandes, and A. Cubillo. doi: 10.5281/zenodo.7753301

Crossref Full Text | Google Scholar

Kioupi, V., and Voulvoulis, N. (2019). Education for sustainable development: a systemic framework for connecting SDGs to educational outcomes. Sustainability 11:601. doi: 10.3390/su11216104

Crossref Full Text | Google Scholar

Laboratorio de Economía de la Educación (LEE) de la Pontificia Universidad Javeriana (2024). Informe no. 92. Pruebas saber 11: Una década de análisis [report no. 92. SABER test 11: A decade of analysis]. Available online at: https://lee.javeriana.edu.co/publicaciones-y-documentos

Google Scholar

Ladrón de Guevara, M., Lopez-Agudo, L. A., Prieto-Latorre, C., and Mercenario-Gutierrez, D. (2022). Internet use and academic performance: an interval approach. Educ. Inf. Technol. 27, 11831–11873. doi: 10.1007/s10639-022-11095-4

Crossref Full Text | Google Scholar

Lee, S.-K., and Fujii, H. (2024a). “Introduction: challenges in science education in the era of sustainable development goals” in Science education for sustainable development in Asia (Singapore: Springer).

Google Scholar

Lee, S.-K., and Fujii, H. (2024b). “Conclusion: toward advancing science education for sustainable development in Asia” in Science education for sustainable development in Asia (Singapore: Springer).

Google Scholar

Liu, X. (2024). Effect of teacher-student relationship on academic engagement: the mediating roles of perceived social support and academic pressure. Front. Psychol. 15, 1–8. doi: 10.3389/fpsyg.2024.1331667

Crossref Full Text | Google Scholar

Lotz-Sisitka, H., and Lupele, J. (2017). “ESD, learning and quality of education in Africa: learning today for tomorrow” in Schooling for sustainable development in Africa (Cham: Springer).

Google Scholar

Meinck, S., and Brese, F. (2019). Trends in gender gaps: using 20 years of evidence from TIMSS. Large-scale Assess. Educ. 7:8. doi: 10.1186/s40536-019-0076-3

Crossref Full Text | Google Scholar

Ministry of Education of Colombia (2006). Estandares básicos de competencias [standards of basic learning outcomes]. Government of Colombia. Available online at: https://www.mineducacion.gov.co/1759/articles-340021_recurso_1.pdf

Google Scholar

Ministry of Education of Colombia (2016a). Derechos básicos de aprendizaje: Ciencias naturales [basic learning rights for science]. Government of Colombia. Available online at: https://www.colombiaaprende.edu.co/sites/default/files/files_public/2022-06/DBA_C.Naturales-min.pdf

Google Scholar

Ministry of Education of Colombia (2016b). Derechos básicos de aprendizaje: Ciencias sociales [basic learning rights for social studies]. Government of Colombia. Available online at: https://www.colombiaaprende.edu.co/sites/default/files/files_public/2022-06/DBA_C.Sociales-V2.pdf

Google Scholar

Morgan, P. L., Hu, E. H., Farkas, G., Hillemeier, M. M., Oh, Y., and Gloski, C. A. (2023). Racial and ethnic disparities in advanced science and mathematics achievement during elementary school. Gift. Child Q. 67, 151–172. doi: 10.1177/00169862221128299

Crossref Full Text | Google Scholar

National Academies of Sciences, Engineering, and Medicine (2023). Closing the opportunity gap for young children. Washington, DC: The National Academies Press.

Google Scholar

O’Flaherty, J., and Liddy, M. (2018). The impact of development education and education for sustainable development interventions: a synthesis of the research. Environ. Educ. Res. 24, 1031–1049. doi: 10.1080/13504622.2017.1392484

Crossref Full Text | Google Scholar

OECD (2023). PISA 2022 results (volume I): The state of learning and equity in education. Paris: PISA, OECD Publishing.

Google Scholar

OECD (2024) How do public and private schools differ in OECD countries? Policy brief. Available online at: https://www.oecd.org/en/publications/how-do-public-and-private-schools-differ-in-oecd-countries_90348307-en.html

Google Scholar

OECD (2025). Education at a glance 2025: Colombia. Available online at: https://www.oecd.org/en/publications/education-at-a-glance-2025_1a3543e2-en/colombia_32e069b7-en.html

Google Scholar

Olsson, D., Gericke, N., and de Boeve-Pauw, J. (2021). The effectiveness of education fo sustainable development revisted – a longitudinal study on secondary students’ action competence for sustainability. Environ. Educ. Res. 28, 405–429. doi: 10.1080/13504622.2022.2033170

Crossref Full Text | Google Scholar

Prangthip, P., Soe, Y. M., and Signar, J. F. (2019). Literature review: nutritional factors influencing academic achievement in school age children. Int. J. Adolesc. Med. Health 33, 1–9. doi: 10.1515/ijamh-2018-0142

Crossref Full Text | Google Scholar

Raudenbush, S., Bryk, A., and Congdon, R. (2019). HLM 8: hierarchical linear and nonlinear modeling. Scientific Software International. Available online at: https://www.hearne.software/getattachment/a104646a-d78e-4ccd-88cf-cf04a78b47f2/HLM8-Manual.aspx

Google Scholar

Reimers, F. (2024). The sustainable development goals and education, achievements and opportunities. Int. J. Educ. Dev. 104:102965. doi: 10.1016/j.ijedudev.2023.102965

Crossref Full Text | Google Scholar

Rodríguez-Gómez, W., Rodríguez, E., and Gomez-Gonzalez, J. (2024). The rural-urban student performance gap in Colombia. Desarro. Soc. 97, 59–75. doi: 10.13043/dys.97.3

Crossref Full Text | Google Scholar

Sachs, J. D., Lafortune, G., and Fuller, G. (2024). The SDGs and the UN Summit of the Future. Sustainable Development Report 2024. Dublin: Dublin University Press. doi: 10.25546/108572

Crossref Full Text | Google Scholar

Sachs, J. D., Schmidt-Traub, G., Mazzucato, M., Messner, D., Nakicenovic, N., Rockström, J., et al. (2019). Six transformations to achieve the sustainable development goals. Nat. Sustain.. 2, 805–814. doi: 10.1038/s41893-019-0352-9

Crossref Full Text | Google Scholar

Schneider, J. (2017). Beyond test scores: A better way to measure school quality. Cambridge, MA: Harvard University Press.

Google Scholar

Sun, Y.-J., Sahakian, B. J., Langley, C., Yang, A., Jiang, Y., Kang, J., et al. (2024). Early-initiated childhood reading for pleasure: associations with better cognitive performance, mental well-being and brain structure in young adolescence. Psychol. Med. 54, 359–373. doi: 10.1017/S0033291723001381

PubMed Abstract | Crossref Full Text | Google Scholar

Tikly, L. (2019). Education for sustainable development in Africa: a critique of regional agendas. Asia Pac. Educ. Rev. 20, 223–237. doi: 10.1007/s12564-019-09600-5

Crossref Full Text | Google Scholar

Uleanya, C. (2025). Quality education versus sustainable development and sustainability in South Africa: scholarly perceptions. J. Furth. High. Educ. 49, 568–584. doi: 10.1080/0309877X.2025.2470763

Crossref Full Text | Google Scholar

UNESCO (2018). Issues and trends in education for sustainable development. Available online at: https://unesdoc.unesco.org/ark:/48223/pf0000261954

Google Scholar

UNESCO (2020). Education for sustainable development: A roadmap. Available online at: https://unesdoc.unesco.org/ark:/48223/pf0000374802

Google Scholar

Vasey, G. (2020). The impact of child labor on student enrollment, effort, and achievement: Evidence from Mexico. Working paper. Department of Economics, University of Pennsylvania. Available online at: https://gabriellevasey.github.io/JMP_GVasey.pdf

Google Scholar

Vásquez-Russi, C. M. (2020). Enseñanzas y aprendizajes sobre la Cátedra de la Paz en Colombia [lessons learned from the implementation of the peace course in Colombia]. Educ. Educ. 23, 221–239. doi: 10.5294/edu.2020.23.2.4

Crossref Full Text | Google Scholar

von Stumm, S., Cave, S. N., and Wakeling, P. (2022). Persistent association between family socioeconomic status and primary school performance in Britain over 95 years. NPJ Sci. Learn. 7:4. doi: 10.1038/s41539-022-00120-3

PubMed Abstract | Crossref Full Text | Google Scholar

Wals, R., and Kief, G. (2010). Education for sustainable development: Research overview. SIDA. Available online at: https://edepot.wur.nl/161396

Google Scholar

Wang, W., Dong, Y., Liu, X., Bai, Y., and Zhang, L. (2020). The effect of parents’ education on the academic and non-cognitive outcomes of their children: evidence from China. Child Youth Serv. Rev. 117:105307. doi: 10.1016/j.childyouth.2020.105307

Crossref Full Text | Google Scholar

World Bank (2018). World development report 2018: Learning to realize education’s promise. Available online at: https://hdl.handle.net/10986/28340

Google Scholar

World Bank (2020). Learning for all: Investing in people’s knowledge and skills to promote development. World Bank Group education strategy 2020. Available online at: https://pubdocs.worldbank.org/en/418511491235420712/education-strategy-4-12-2011.pdf

Google Scholar

World Bank, UNESCO, UNICEF (2021). The state of the global education crisis: A path to recovery. Available online at: https://www.unicef.org/reports/state-global-education-crisis

Google Scholar