Geographical differences in Italian student's English test performance: a role of individual and local characteristics

Several studies have been focused on the analysis of the factors that affect the school performance. It has been found that social, economic and environmental factors have an important role in determining the educational achievement, and that individual characteristics and family background affect the school performance more than school characteristics. The spatial statistical analysis has been used by many researchers to study educational achievement. Principally, these studies concentrated their attention on the relationship between school performance and socio-economic variables of the catchment areas from the spatial perspective, by using Geographically Weighted Regression (GWR). The aim of this study is to investigate which individual characteristics may affect test scores in English Reading and Listening and which local characteristics of the catchment areas influence students’ achievement in these tasks. The data we used come from the National INVALSI Assessment of 8th grade for the 2018/2019 academic year and includes results of the standardized test in English Reading and Listening. Firstly, through the simple correlation analysis, we defined the individual and local characteristics that were then used as the explanatory variables in our analysis. Secondly, the simple Ordinary Least Squares (OLS) regression was used to analyze the relation of students’ performance in both tasks to previously selected characteristics. To verify a presence of positive spatial autocorrelation within the Italian territory, we calculated a widely used spatial statistic Global Moran’s I. It helped us to estimate the strength of spatial correlation, and to test also the significance of the spatial correlation. The positive result of Moran’s I endorsed our intention to include the territorial dimension into our study, and therefore, the GWR was used to model spatially varying relationships between the students’ achievement and the explanatory variables tested with the OLS regression. However, the emphasis here is not so much to determine whether or not exist relationships between school performance and catchment area characteristics, but to determine if there are any interesting spatial variations in these relationships.