Abstract
This study addresses the scarcity of quantitative assessments for elementary education in China by leveraging big data from English academic proficiency tests (2018–2020) across 322 senior high schools in Gansu Province. By developing the "English Excellence Degree" (EED) metric and applying spatial-temporal analysis methods, the research identifies significant disparities in educational quality. Findings reveal a staggering 759-fold difference between the highest and lowest performing schools, with excellence concentrated in the provincial capital, Lanzhou, and "cold spots" prevalent in minority and impoverished regions. These results underscore the utility of big data in guiding regional education policy and resource allocation.
Introduction
In the era of digital transformation, big data has become a pivotal tool for industries and governments. However, its application in assessing elementary education levels—particularly in developing regions—remains limited. Gansu Province, located in northwest China, represents an economically challenged area where English education quality varies significantly. Traditionally, educational assessments have relied on subjective generalizations. This study shifts the paradigm by using large-scale test data to scientifically map educational inequality and provide a foundation for balanced regional development.
Methodology
The researchers analyzed score data from 322 senior high schools over a three-year period. Central to the study is the English Excellence Degree (EED), a composite indicator calculated as:
$$EED = M_j \times E_j$$
where $M_j$ is the school's average English score and $E_j$ is its passing rate. This metric offers a more comprehensive view of performance than raw scores alone.
To analyze the spatial distribution of these scores, the team employed:
- Anomaly Analysis: To measure deviations from the regional mean.
- Kernel Density Estimation (KDE): To visualize the concentration of high-performing schools.
- Getis-Ord $G_i^*$ Statistics: To identify statistically significant "hot spots" (high-performance clusters) and "cold spots" (low-performance clusters).
- Cluster Analysis: To categorize schools into four distinct performance groups based on similarity.
Results and Spatial Patterns
The analysis uncovered extreme educational stratification within the province:
- The Excellence Gap: The highest average EED recorded was 759 times greater than the lowest, highlighting a massive divide in educational outcomes.
- Negative Anomalies: More than 50% of the surveyed schools performed below the regional average, indicating that a majority of institutions struggle to meet baseline quality standards.
- The "Lanzhou Effect": Spatial analysis confirmed that Lanzhou, the provincial capital, serves as the primary hot spot for English education. Conversely, cold spots are heavily concentrated in the Linxia autonomous region, Longnan, and eastern Qingyang—areas often characterized by higher poverty rates or minority populations.
- Temporal Trends: While violin plot analysis showed a general upward trend in EED values from 2018 to 2020, the progress was uneven; 136 schools actually saw a decrease in quality during this period.
Discussion: The Teacher Factor
To understand the drivers behind these data points, the researchers conducted questionnaires. The qualitative findings suggest that the "cold spots" are trapped in a cycle of disadvantage. High-quality teachers overwhelmingly prefer urban centers with better living conditions and professional opportunities. This "teacher flight" leaves rural and minority-area schools with fewer resources, further depressing student performance and widening the regional gap.
Conclusion
This case study demonstrates that big data is an essential tool for the quantitative assessment of basic education. By pinpointing specific geographic areas and schools that are falling behind, policymakers can move beyond anecdotal evidence to implement targeted interventions. Future research will expand this framework to include micro-level factors, such as individual student study behaviors and the specific impact of teacher rotation policies, to further refine strategies for educational equity.
Summary updated