Using Predictive Analytics in the Real World: Practice Scenarios for Ethical Decision‐Making

Autism & Developmental Language Impairments

Alverson

2022

Background and Aims The fastest growing group of students with disabilities are those with Autism Spectrum Disorder (ASD). States annually report on post-high school outcomes (PSO) of exited students. This study sought to fill two gaps in the literature related to PSO for exited high-school students with ASD and the use of state data and predictive modeling. Methods Data from two states were analyzed using two predictive analytics (PA) methods: multilevel logistic regression and machine learning. The receiver operating characteristic curve (ROC) analysis was used to assess predictive performance. Results Data analyses produced two results. One, the strongest predictor of PSO for exited students with ASD was graduating from high school. Two, machine learning performed better than multilevel logistic regression in predicting PSO engagement across the two states. Conclusion This study contributed two new and important findings to the literature: (a) PA models should be applied to state PSO data because they produce useful information, and (b) PA models are accurate and reliable over time. Implications These findings can be used to support state and local educators to make decisions about policies, programs, and practices for exited high school students with ASD, to help them successfully transition to adult life.

Section: Discussionmentioning

confidence: 99%

From high school to postsecondary education, training, and employment: Predicting outcomes for young adults with autism spectrum disorder

Autism & Developmental Language Impairments

Alverson

2022

“…There is agreement across most fields that personal human judgment remains central, and all of these PA tools should only be used in support of, and not in place of, that judgment (Chin-Yee and Upshur 2017). As analytic tools continue to advance, ethical and equitable applications of PA to state administrative data (see Palmer and Carpenter-Hubin 2020) must always be ensured to move the field forward for everyone.…”

Section: Discussionmentioning

confidence: 99%

Post-High School Outcomes of Adolescents with Learning Disabilities: Using Annual State Administrative Data and Predictive Analytics

Journal of Applied Social Science

2022

This study involved the analyses of extant data from two U.S. states of post-school outcomes (PSO) for students with a specific learning disability (SLD) one year after they had exited high school. The purpose of this study was to fill two gaps in the literature. The first gap was to understand what happened to these exiters in the first year after high school related to employment and further education or training at a state level. The second gap was to demonstrate the necessity of local and state education professionals to use PSO data, which is collected annually, by applying predictive analytics (PA) to support their decision making. The data analyses produced two main findings. One, the strongest predictors of PSO were students graduating from high school and their high school classroom placement. Two, PA was reasonably accurate in predicting PSO and demonstrated robust capabilities for reliable use on an annual basis to support policies, programs, and practices. Limitations of this study related to the data and number of predictors. The study concludes with implications of administrative state data use and PA for state and local education professionals and for researchers.

“…There is general agreement across research fields that human judgment remains central, and that these analytic tools should only be used in support of that judgment, not in its place (Chin-Yee & Upshur, 2017). As technology continues to advance, ethical and equitable applications of PA to data (see Palmer & Carpenter-Hubin, 2020) must always be ensured to move these fields forward for everyone.…”

Section: Machine Learningmentioning

confidence: 99%

Outcomes of Students With Disabilities After Exiting From High School: A Study of Education Data Use and Predictive Analytics

Alverson

2022

Journal of Education

We conducted a study of predictive analytics (PA) applied to state data on post-school outcomes (PSO) of exited high-school students with disabilities (SWD). Data analyses with machine learning Random Forest algorithm and multilevel Bayesian ordered logistic regression produced two key findings. One, Random Forest models were accurate in predicting PSO. Two, Bayesian models found high-school graduation was the strongest predictor of higher education and reliably predicted the specific type of outcome relative to other outcomes. Limitations of this study are the data source and small number of predictors. Implications of the study for researchers and educators are discussed in conclusion.