Recent analyses indicate that educational achievement gaps based on student socio-economic status (SES) have grown over recent decades, that is, SES-achievement association has become stronger (Chmielevski, 2019; Harwell et al., 2017). Current gaps in achievement are attributed to socio-economic differences among student families rather than a lack of access to or selectivity inside the schooling systems (Chmielevski, 2019). From this perspective, student SES is considered a students’ personal background characteristic and treated as a factor that can hardly be changed or influenced by educational policies or practices (Holzberger et al., 2020; Zabulionis, 2020). However, it is also acknowledged that SES can be related to student academic achievement through the institutional level. Specifically, when considered on a cumulative level (i.e., aggregated by school or classroom), student SES becomes a characteristic of school or educational system. It is referred to in the literature as aggregate-level SES, school SES or school composition (Van Ewijk & Sleegers, 2010). Aggregate-level SES is related to multiple factors and processes at school, including educational resources, school climate, instructional quantity and quality (Gustafsson et al., 2018), as well as school practices related to sorting of students (Van Ewijk & Sleegers, 2010).

There is an agreement in the literature that both individual and school or classroom-level effects of SES are important for explaining achievement, but there is a substantial variation of correlation patterns among different countries (Yang & Gustafsson, 2004). For example, an overview of studies on the relationship between aggregate-level SES and achievement suggests that the effect varies from null to strong (Holzberger et al., 2020; Van Ewijk & Sleegers, 2010). Moreover, aggregate-level school SES was shown to moderate the links between student-level SES and academic achievement, but the effect was of opposite direction across different countries (Gustafsson et al., 2018). Overall, these findings point to the need for a deeply contextualized understanding of the SES-achievement link on both - individual and institutional level.

Considering this, the aim of our study is to further the understanding of different aspects of SES-achievement association by analyzing nationally representative data for Lithuania, an EU country for which there are few findings from multi-level analyses on SES-achievement link. Over the recent decades, socio-economic inequalities have sharpened in the Lithuanian society and could be considered among the highest in the EU (Lazutka, Juška, & Navickė, 2018). Moreover, social justice disparities in the field of secondary education were observed across the rural-urban divide inside the country. Specifically, the existing findings indicate that rural residents have more limited possibilities to participate in higher secondary education compared to urban residents (Žalimienė et al., 2011), however, such findings are over a decade old, which points to the need for a more current perspective on rural-urban educational disparities. Therefore, our first objective is to estimate the extent to which socio-economic inequalities in Lithuania are reflected in academic achievement on a student and school level, taking into account potential variations across the rural and urban context.  

In addition, we consider the previous finding that Lithuania is among the countries where educational system is compensatory with respect to student SES (Gustafsson et al., 2018), i.e., in schools with a high aggregate-level SES, the relationship between SES and achievement on a student level is weaker than in low-SES schools. This insight comes from TIMSS 2011 assessment of 8^th grade students. However, there is no information if such cross-level interaction could replicate in other available country-wide datasets and cohorts in Lithuania. Thus, our second objective is to see whether the interaction between student and school-level SES is observed in more recent nationally collected data.

In this study we analysed nationally representative data from the National Student Achievement Survey 2014 cohort of Lithuanian 8th grade students (N = 3763; 50.5% females). The data was collected by the National Examination Center and provided in an open-access format. The participants came from 178 classes in 148 schools. We used school for aggregate-level analyses. There were 7 to 68 participants within one school. One fifth of the participants came from rural residential areas, the rest came from medium-sized (47.9%) or large (32.4%) cities. Student SES was measured with a composite index, which covered a relatively broad range of home material and educational resources, including books, computers, art pieces, separate room, home appliances, and pocket money per week. The items were added to form an overall index and the scale was transformed to have a mean of 0.50 and a standard deviation of 0.10. Three cases were removed due to missing data on SES variable. Aggregate-level SES was defined within a multi-level framework as the mean level of SES of the students in a school (ICC = .15). Academic achievement was measured in three subject areas: math, reading (Lithuanian), and social sciences. Each student received standardised tests in one or two subject areas; the number of responses for each subject area ranged from 1257 for social sciences to 1702 for math. The scale was transformed to have a mean of 500 and a standard deviation of 100. Differences between schools accounted for a substantial share of variance in achievement on all three subjects (ICCmath = .16; ICCread = .17; ICCsoc = .18). For our main analyses we used multilevel regression (MLR) with Mplus 8.4. First, we sought to understand how SES predicted academic achievement at student and school levels and investigated the presence of school composition effects. Next, we applied multigroup analyses to find potential variations across two residential contexts (urban and rural schools). Finally, we sought to understand if the effects of SES on academic achievement varied across schools, and if they were predicted by aggregate-level SES, so we tested a series of random-slope MLR models with cross-level interactions.

We found weak effects of student SES on achievement scores at the individual level and these weak student-level effects were similar across urban and rural contexts. We found very strong effects of school-level SES on academic achievement and these effects were substantially higher in urban context (share of variance explained: .58 for math and .53 for social sciences) than in rural context (share of variance explained: .11 for math and .18 for social sciences), except for achievement in reading (.46 and .50, respectively). Moreover, we found a presence of compositional effects, indicating that school-level SES predicted achievement over and above the individual-level SES differences. These compositional effects were stronger in urban context compared to rural context. Finally, our analyses of cross-level interactions produced a number of convergence issues and were thus inconclusive. We discuss our findings and their implications in a comparative and national educational context.

Chmielewski, A. K. (2019). The global increase in the socioeconomic achievement gap, 1964 to 2015. American Sociological Review, 84(3), 517–544. https://doi.org/10.1177/0003122419847165 Gustafsson, J.-E., Nilsen, T., Yang Hansen, K. (2018). School characteristics moderating the relation between student socio-economic status and mathematics achievement in grade 8. Evidence from 50 countries in TIMSS 2011. Studies in Educational Evaluation, 57, 16-30. https://doi.org/10.1016/j.stueduc.2016.09.004 Harwell, M., Maeda, Y., Bishop, K., & Xie, A. (2017). The surprisingly modest relationship between SES and educational achievement. The Journal of Experimental Education, 85(2), 197–214. https://doi.org/10.1080/00220973.2015.1123668 Holzberger, D., Reinhold, S., Lüdtke, O., & Seidel, T. (2020). A metaanalysis on the relationship between school characteristics and student outcomes in science and maths – evidence from large-scale studies. Studies in Science Education, 56(1), 1-34. https://doi.org/10.1080/03057267.2020.1735758 Lazutka, R., Juška, A., & Navickė, J. (2018). Labour and capital under a neoliberal economic model: Economic growth and demographic crisis in Lithuania. Europe-Asia Studies, 70(9), 1433–1449. https://doi.org/10.1080/09668136.2018.1525339 Van Ewijk, R., & Sleegers, P. (2010). The effect of peer socioeconomic status on student achievement: A meta-analysis. Educational Research Review, 5, 134–150. https://doi.org/10.1016/j.edurev.2010.02.001 Yang, Y., & Gustafsson, J.-E. (2004). Measuring socioeconomic status at individual and collective Levels. Educational Research and Evaluation, 10(3), 259–288. https://doi.org/10.1076/edre.10.3.259.30268 Zabulionis, A. (2020). Tarptautinio švietimo tyrimo OECD PISA Lietuvos ir kaimyninių šalių duomenų tikslinė antrinė analizė. ŠMSM, NŠA, Vilnius. Žalimienė, L., Lazutka, R., Skučienė, D., Aidukaitė, J., Kazakevičiūtė, J., Navickė, J., & Ivaškaitė-Tamošiūnė, V. (2011). Socialinis teisingumas švietime: Teorinė samprata ir praktinis vertinimas. ŠMM, LSTC, Vilnius.