Mode of Program Participation
Academic Scholarship
Participation Type
Paper
Presentation #1 Title
A Data-Based Discriminant Model to Predict Educational Attainment in Central Appalachia
Presentation #1 Abstract or Summary
The purpose of this study was to develop a data-based model to predict case classification of central Appalachian students’ educational attainment levels into one of three categories: will not enroll in post-secondary; will enroll in, but not complete a bachelor's degree; and will enroll in and complete a bachelor’s degree. The research question that guided this study was: What predictors of educational attainment significantly contribute to a discriminant function capable of classifying central Appalachian students into one of three educational attainment groups? This research was grounded in the work of Sewell and Hauser (1972). Antecedents to educational attainment for this current study were selected based on their work. Restricted-use data were obtained from the National Center for Education Statistics, Education Longitudinal Study of 2002. This study was bound by counties designated by the Appalachian Regional Commission as being part of central Appalachia, north-central Appalachia, and south- central Appalachia, collectively referred to as central Appalachia in this study. Step-wise discriminant analysis was used to develop a model to predict likely educational attainment outcomes for central Appalachian students eight years into the future. By collecting three pieces of information from the student, and verifying their GPA, likely educational outcomes were predicted with 65.8% accuracy in this study using the leave-one-out method. This paper presents the results of this work to develop an educational attainment model for central Appalachian students. This session will include future research plans to further test and vet this model for use by practitioners.
At-A-Glance Bio- Presenter #1
Ferris is an assistant professor and MBA program director at the University of Charleston, in Charleston WV where he lives and has worked full-time for four years. Prior to that he spent four years as a United States Marine, 10 years as an IT Manager for United Way, about 10 years in sales of, and consulting about, IT related products and services, and about 11 years of part time teaching experience. He has taught over 80 college sections in Marketing, Management, Technology, and Leadership.
A Data-Based Discriminant Model to Predict Educational Attainment in Central Appalachia
The purpose of this study was to develop a data-based model to predict case classification of central Appalachian students’ educational attainment levels into one of three categories: will not enroll in post-secondary; will enroll in, but not complete a bachelor's degree; and will enroll in and complete a bachelor’s degree. The research question that guided this study was: What predictors of educational attainment significantly contribute to a discriminant function capable of classifying central Appalachian students into one of three educational attainment groups? This research was grounded in the work of Sewell and Hauser (1972). Antecedents to educational attainment for this current study were selected based on their work. Restricted-use data were obtained from the National Center for Education Statistics, Education Longitudinal Study of 2002. This study was bound by counties designated by the Appalachian Regional Commission as being part of central Appalachia, north-central Appalachia, and south- central Appalachia, collectively referred to as central Appalachia in this study. Step-wise discriminant analysis was used to develop a model to predict likely educational attainment outcomes for central Appalachian students eight years into the future. By collecting three pieces of information from the student, and verifying their GPA, likely educational outcomes were predicted with 65.8% accuracy in this study using the leave-one-out method. This paper presents the results of this work to develop an educational attainment model for central Appalachian students. This session will include future research plans to further test and vet this model for use by practitioners.
