Developing a Hybrid Model to Predict Student First Year Retention in STEM Disciplines Using Machine Learning Techniques.

1. IntroductionUnderstanding the reasons behind the low enrollment and retention rates of Underrepresented Minority (URM) students (African Americans, Hispanic Americans, and Native Americans) In the disciplines of science, technology, engineering, and mathematics (STEM) has concerned researchers for decades. Statistics show that students of color have higher attrition rates compared with other groups, although this trend has been decreasing over the past twenty years (Besterfield-Sacre, Atman, & Shuman, 1997; Mitchell & Daniel, 2007; Fleming, Ledbetter, Williams, & McCain, 2008). These groups tend to enroll in STEM majors in small numbers and leave in higher numbers (Urban, Reyes, & Anderson-Rowland, 2002; Alkasawneh & Hobson, 2009).Increasing the number of minorities (women and ethnic groups) Is a practical way of increasing the workforce pool in STEM fields where white male representation is still dominant. Unfortunately, this solution is difficult for many Institutions. Only two out of five African American and/or Hispanic American students remain In their majors and receive bachelor's degrees in a STEM discipline nationwide (Markley, 2005). In order to impact workforce demographics, the population of students choosing STEM majors must change. The literature reflects a substantial Interest in increasing URM student retention In higher education (Sidle & McReynolds, 1999; Nave, Frizell, Obiomon, Cui, & Perkins, 2006; Hargrove & Burge, 2002). Retention Is ofslgnlficant Interest because of Its positive impact on college reputation and workforce demographics (Williford & Schaller, 2005).Several studies emphasize the importance of Identifying college students with higher risk of dropping out in early stages In order to allocate the available resources based upon student needs (Herzog, 2006; Lin, Imbrie, & Reid, 2009). Research by Zhang, Anderson, Ohland, Carter, &Thorndyke (2002) stated that identifying factors that affect student retention could play an effective role in the counseling and advising process for engineering students. This equips Institutions to utilize their available resources based upon those groups' needs (Herzog, 2006). Traditional methods of statistical analysis have been used to predict student retention, such as logistic regression (Gaskins, 2009). Recently, research has focused on data mining techniques to study student retention In higher education (Brown, 2007). These techniques are highly accurate, robust with missing data, and do not need to be built on a hypothesis. Data mining Is defined as recognizing patterns in a large set of data and then trying to understand those patterns.1.1 Predictive models of student retention7.7.7 Tinto's modelTinto In his model (1975) noted that integration Into the college system, academically and socially, impacts students'decision regarding dropping out of college. He added that integration into the college system causes a continuous change in student goals and commitment to graduation, which In turn might generate the decision of persistence or dropping out of college. Tinto's model was based on Durkheim's theory of suicide (Durkheim, 1951) which clearly connected suicide rates to individuals' Integration In the community.Variables included in this model are Individual attributes such as gender and race, pre-college experiences, and family backgrounds. Tinto argues that these variables influence the development of college expectations and commitment to graduation. These expectations and commitments are modified based upon Integration into the college system academically and socially to generate a new level of commitment and goals.The author noted thatthere Is still little Information that links race with college dropouts, although it is considered a strong predictor of student persistence. Tinto further added that there isn't enough knowledge about the process of interaction that leads racial groups to drop out and how these processes are affecting their academic and social integration (Tinto, 1975). …

[1]  John Heywood,et al.  Engineering Education : Research and Development in Curriculum and Instruction , 2005 .

[2]  Judy A. Perkins,et al.  Prairie View A&M University: Assessing the Impact of the STEM-Enrichment Program on Women of Color , 2006 .

[3]  E. Durkheim.  Suicide: A Study in Sociology , 1897 .

[4]  Ruba Alkhasawneh,et al.  Summer Transition Program: A Model For Impacting First Year Retention Rates For Underrepresented Groups , 2009 .

[5]  Mykola Pechenizkiy,et al.  Predicting Students Drop Out: A Case Study , 2009, EDM.

[6]  Deborah Mayhew,et al.  Evaluation Of Retention And Other Benefits Of A Fifteen Year Residential Bridge Program For Underrepresented Engineering Students , 2009 .

[7]  Serge Herzog,et al.  Estimating Student Retention and Degree-Completion Time: Decision Trees and Neural Networks Vis-a-Vis Regression. , 2006 .

[8]  Johan Malmqvist,et al.  Rethinking Engineering Education - The CDIO Approach , 2007 .

[9]  Jason Brown,et al.  Neural Network Prediction of Math and Reading Proficiency as Reported in the Educational Longitudinal Study 2002 Based on Non-Curricular Variables , 2007 .

[10]  Meg Wright Sidle,et al.  The Freshman Year Experience: Student Retention and Student Success , 1999 .

[11]  Catherine C. Marshall,et al.  Designing Qualitative Research , 1996 .

[12]  Amy Freeman,et al.  Creating a Successful Model for Minority Students' Success in Engineering: The PREF Summer Bridge Program , 2005 .

[13]  Brady P Gaskins.  A ten-year study of the conditional effects on student success in the first year of college , 2009 .

[14]  Lorraine Fleming,et al.  Engineering Students Define Diversity: An Uncommon Thread , 2008 .

[15]  Nadine Meskens,et al.  Determination of factors influencing the achievement of the first-year university students using data mining methods , 2006 .

[16]  Vincent Tinto.  Dropout from Higher Education: A Theoretical Synthesis of Recent Research , 1975 .

[17]  A. Williford,et al.  All Retention all the Time: How Institutional Research Can Synthesize Information and Influence Retention Practices , 2005 .

[18]  R. Reason.  Student Variables that Predict Retention: Recent Research and New Developments , 2003 .

[19]  J. J. Lin,et al.  Student Retention Modelling : An Evaluation of Different Methods and their Impact on Prediction Results , 2009 .

[20]  Ernest T. Pascarella,et al.  Toward the validation of Tinto's model of college student attrition: A review of recent studies , 1980 .

[21]  M.R. Anderson-Rowland.  A first year engineering student survey to assist recruitment and retention , 1996, Technology-Based Re-Engineering Engineering Education Proceedings of Frontiers in Education FIE'96 26th Annual Conference.

[22]  Daryl E. Chubin,et al.  A Retrospective on Undergraduate Engineering Success for Underrepresented Minority Students , 2003 .

[23]  A. Astin,et al.  Assessment for Excellence: The Philosophy and Practice of Assessment and Evaluation in Higher Education. American Council on Education/Macmillan Series on Higher Education. , 1991 .

[24]  S. K. Hargrove,et al.  Developing a six sigma methodology for improving retention in engineering education , 2002, 32nd Annual Frontiers in Education.

[25]  Jody Markley,et al.  Integrating a Faculty Directed Research Experiences into a High School Bridge Program , 2005 .

[26]  Vincent Tinto,et al.  Stages of Student Departure: Reflections on the Longitudinal Character of Student Leaving , 1988 .

[27]  Paul C. Lam,et al.  Predicting Success in a Minority Engineering Program , 1999 .

[28]  Guillermo Mendez,et al.  Factors Associated With Persistence in Science and Engineering Majors: An Exploratory Study Using Classification Trees and Random Forests , 2008 .

[29]  A. Astin.  Student involvement: A developmental theory for higher education. , 1999 .

[30]  James Mathias,et al.  A short-term assessment of a multi-faceted engineering retention program , 2009, 2009 39th IEEE Frontiers in Education Conference.

[31]  Nguyen Thai Nghe,et al.  A comparative analysis of techniques for predicting academic performance , 2007, 2007 37th Annual Frontiers In Education Conference - Global Engineering: Knowledge Without Borders, Opportunities Without Passports.

[32]  Cynthia J. Atman,et al.  Characteristics of Freshman Engineering Students: Models for Determining Student Attrition in Engineering , 1997 .