Small Changes in School Interactions Can Reverse Segregation

Students with special education needs (SEN) often face discrimination and segregation in schools, despite international policies promoting inclusive education. This persistent gap between policy and practice stems from factors like untrained educators and a public perception that inclusion requires minimal effort, leading to self-perpetuating patterns of segregation. A new theoretical study uses computational modeling to demonstrate that small variations in how students perceive and interact with each other can reverse these trends, generating stable patterns of inclusivity through empathy propagation.

The researchers found that by adjusting simple rules in a model of social interaction, they could shift outcomes from segregation to inclusion. The study identified four scenarios based on variations in two key parameters: SENstudents and nonSENstudents, which represent the influence of students with and without special education needs. In the baseline scenario (Scenario 1), with both parameters set to 4, segregation patterns emerged, mirroring the classic Schelling model where students self-segregate without a particular reason. However, when these parameters were slightly modified—such as increasing SENstudents to 5 or nonSENstudents to 5—the model produced scenarios where empathy and inclusivity became dominant. Specifically, Scenario 4, with both parameters set to 5, showed the most effective convergence toward inclusive patterns, where students' opinions shifted toward perceiving others as equals.

Ology relied on cellular automata, a type of computational model that simulates social dynamics through simple rules applied to a grid of cells representing students. Each cell was assigned a state of SEN (value 1) or nonSEN (value 0), and opinions could change based on the number of neighboring cells with similar or dissimilar states. The researchers used a Moore neighborhood, which considers all eight surrounding cells, to simulate real-world social influences. They tested different initial proportions of SEN and nonSEN students—such as 50-50, 60-40, and 40-60 splits—and ran simulations over 15 iterations to observe how patterns evolved. To validate their , they compared the deterministic scenarios against null models based on binomial distributions with probabilities of 0.3, 0.5, and 0.7 for selecting SEN students, ensuring were statistically significant and not due to random chance.

, Detailed in Figures 4, 5, and 6, show clear visual and quantitative evidence of the shift from segregation to inclusion. In Figure 4, Scenario 1 displayed the classical segregation pattern, with a stable percentage of nonSEN students around 50%, while Scenario 4 showed a rapid decline in nonSEN percentages, indicating widespread empathy propagation. Figure 5 illustrated these patterns in grid structures, where Scenario 4 achieved complete convergence of empathy by Time = 10. Table 1 confirmed that all scenarios were statistically different from the null models, with correlation coefficients ranging from weak to moderate and often negative, indicating that the deterministic rules drove the outcomes. Additionally, Figure 6 explored how initial conditions affected ; for example, a higher initial proportion of SEN students (70%) accelerated inclusive patterns, while a higher proportion of nonSEN students (70%) slowed but did not prevent empathy dissemination in the best-case scenario.

Of these are profound for real-world education systems. The study suggests that inclusive patterns can emerge from small, local changes in student interactions, such as slightly increasing peer pressure from SEN students or enhancing resistance to negative influences from nonSEN students. This s the notion that segregation is inevitable and offers a theoretical basis for interventions that promote empathy in schools. The researchers provided specific recommendations, including school board meetings to share inclusive education information, parent organizations reinforcing empathetic behaviors, and coalitions between families and educators to monitor progress. They also proposed public policies, such as allocating 7.5% to 9.5% of GDP to education with a focus on high-impact areas like teacher training and early childhood interventions.

However, the study acknowledges several limitations. The gap between educational policies and local school issues remains a significant barrier, as the theoretical rules may not translate to real-world success without coordination among students, families, educators, and sectors. The model assumes an idealized environment with no physical or social barriers, using a torus configuration where edges are connected, which may not reflect the complexities of actual classrooms. Additionally, the study's focus on theoretical modeling means that practical implementation requires concrete action plans and government support, including effective fund allocation and legal frameworks. Future work should address these s by designing public policies that permeate the entire educational system, particularly through early childhood investments that yield long-term economic and social returns.