Predicting Appropriate Educational Environments for Special Education

No explicit and objective method using historical national data sets is universally implemented in United States (US) public schools to predict the appropriate educational environments for students in special education (SPED) . How are national K-12 special education data sets of students’ background characteristics, academic proficiency, behavior, and needs related to trends in educational environments? Categorization is a causal-model theory, which explains the potential for developing an explicit and objective method for using historical national data to predict appropriate special education environments for students nationwide. Based on a modeling of previous work examining relationships between student background characteristics and STEM pathways (Gottlieb, 2018), this study will use a logistic regression model to identify factors related to student educational environments in special education, to explore how those factors change as the educational environment changes from one of three selected placements, and to examine how these factors vary with educational environments across the identified student and district characteristics. Casual-Model Theory showed consistent connections to previous literature to highlight the different weights of inputs for decisions about educational environments for students with disabilities. The concept of Category Stretching (Durand & Paolella, 2013) also aligned in that the weight of the student characteristics and district factors influencing placement decisions can change over time due to changes in state policies. The conceptual idea of LRE also consistently remained aligned with the methods and findings of this research, exhibiting how states approach keeping kids with disabilities in the regular classroom to the maximum extent appropriate. As state policies change, it is plausible that, due to category stretching, the casual power of the many independent variables impacting placement decisions may fluctuate. However, based on the data from the USED on the 2018-2019 school year, and the results of this logistic regression, the independent variables of age, special education funding per pupil, reading proficiency, discipline, and disability have the greatest casual weight toward educational environments in special education.