Geospatial analyses to determine academic success factors in California’s K-12 education

Abstract

Standardized test scores are often used to measure students' academic success. Although factors that affect student success involve teaching techniques, classroom dynamics, and study skills, there are other factors outside the classroom that could influence students' overall academic performance. Oftentimes, these factors are overlooked or easily deemed uncontrollable by educators. Prior studies have identified and examined such factors; however, for this analysis, we will use Geographic Information Systems (GIS) to analyse and display spatial patterns of these external factors, such as household income and average household size, which were not previously possible. Utilizing GIS and variations of demographic and lifestyle data allows us to take a closer look into understanding the factors that positively or negatively correlate with academic achievement. We use a sample of 2015-2016 Scholastic Assessment Test (SAT) scores from the California Department of Education and location-based household spending, socioeconomic, and demographic data from Environmental Systems Research Institute, also known as Esri, to develop statistical models in order to understand factors that influence SAT scores. Our results indicate that two parent households and spending on health insurance have a positive effect on student academic achievement. In addition, students that are surrounded by educational businesses score higher on the SAT. We also learn that diversity, household size, and multigenerational households have negative impacts on students' SAT scores.

Shifting spatial patterns in German population trends: local-level hot and cold spots, 1990–2019

“…An applied sensitivity analysis has shown that the fixed-distance-band method provides the most differentiated spatial picture, allowing all spatial clusters of both the hot-and the cold-spot analysis to appear. This method allows the best-fitting interpretation of traditional mappings of population change rates, is a good option for polygon data when there is a large variation in polygon size ensuring a consistent scale of analysis, and does not apply a deterministic assumption about neighborhood relationships or distances (Chew et al, 2020;Kim and Choi, 2017;Sanchez-Cuervo and Aide, 2013;Sánchez-Martín et al, 2019).…”