Online Student Retention

I had the opportunity to present a paper this week at the Association for the Advancement of Computing in Education’s (AACE) E-Learn 2010 conference in Orlando along with Dr. Phil Ice, our Director of Course Design, Research & Development.  The paper, Comprehensive Assessment of Student Retention in Online Learning Environments, originated from research that I conducted as part of my doctoral dissertation at The University of Pennsylvania’s Graduate School of Education.

Student retention has been an issue in higher education since the late 1800’s.  Some of the early research in the area began in the 1930’s but the volume of research studies increased substantially in the 1960’s through the present era.  Early research focused on psychological reasons for students to drop out of college but most of the literature since the late 1970’s have focused on sociological issues.

While there are many significant contributors to the research of student retention (William Spady, Alexander AstinVincent Tinto, John M. Braxton, George Kuh, Ernest Pascarella, etc.), Vincent Tinto’s Leaving College: Rethinking the Causes and Cures of Student Attrition provides a fairly thorough overview of many research studies (note:  the second edition of this book was published in 1994 and new copies are difficult to find).

Tinto is given credit for developing the first theoretical construct of predicting student retention.  Three major areas of his theory involve the importance of the background characteristics of college students, the social integration of students with their college, and the academic integration of students with their professors and programs.  Subsequent research studies have focused on some of these areas and/or attempted to prove or disprove the original construct.  Colleges and universities use some of the studies as the basis for their internal retention research or external explanations of their graduation rate.

Continually, studies contribute to the theory that background characteristics of students are most influential for their persistence in college.  Among the data included in background characteristics studies are SAT and ACT scores, high school Grade Point Average, educational background of parents, and socioeconomic status of the student’s family.  Students with above average profiles in all of these areas usually graduate from college while students with profiles below the averages are less likely to attend or graduate from college.

At American Military University (AMU) and American Public University (APU), our students are primarily working adults with an average age of 30.  While there were notable studies of student retention at adult-serving institutions, I found very few that studied persistence at online universities or studied students who completed their programs entirely online.  While we knew the relative percentage of our students who graduate, we had not analyzed our student database for characteristics that may be significant indicators for a student’s graduation success.

Whenever institutions examine student retention, the analysis has to include the outcomes for students over a certain period of time.  Given that graduate students have already obtained an undergraduate degree, we eliminated them from the analysis.  We also needed to cover a reasonable period of time that would allow students to graduate, remain active, or disenroll from school.  We decided to examine statistics from approximately 20,500 students who were pursuing an undergraduate degree and who completed at least one course in 2007.  The students were grouped into three groups based on their academic standing as of December 31, 2009.  Those groupings were:  students who graduated, students who were still actively taking at least one course per year, and students who had disenrolled as of the end of 2009.

The analysis was fruitful in that it provided me with results that I could convey to our executive team for future studies and continued analysis.  One important finding was that the average length of time it took a student from that group to earn a bachelor’s degree at AMU or APU was 6.7 years.  We already knew that most of our students were engaged full-time with their military or civilian careers, so the length of time to complete a four-year degree would take longer than someone who was enrolled as a full-time student.  However, more than 85 percent of our students transfer in academic credits from a previously attended institution or from workplace training where the learning had been evaluated for academic credit.  Factoring in the average number of credit hours transferred in to our institution with the average number of courses taken in a year, the study results indicate that many successful adult students complete their academic journey over a period of time approaching 10 years.  That 10-year period may be longer if career or family obligations interrupt the flow of taking classes.

We utilized a forward regression model to determine if there were independent variables that might predict whether an individual student would disenroll from either AMU or APU.   It was not all too surprising that students from the group who transferred no credits were most likely to disenroll (our paper was listed as 31084 and will be available as part of the Printed Proceedings book available at  Given that 121 credit hours (40 courses) are required for most bachelor’s programs and the average student at AMU/APU takes 12 credits (4 courses) per year, a 10-year commitment may be easier to walk away from in the early stages than after a substantial amount of credits have been earned.  Supporting this regression analysis outcome is the fact that 40.3 percent of the students in the group who disenrolled did so after taking only 2 classes and 65.3 percent disenrolled after taking 4 or fewer classes.  Decreasing the percentage of students who disenroll early has been a focus at American Public University System (APUS) for nearly a decade and continues.  Many on-ground programs have similar issues with new students leaving early and have programs dedicated to the first year student experience in order to improve the student experience and more fully integrate them into the social and academic culture.  Social and academic integration in online learning environments is more difficult, but not impossible and we will continue to investigate ways in which we can meet the needs of new students in their first classes.

Academic institutions like APUS that participate in the Federal Student Aid program are required to submit vast quantities of student data to the National Center for Education Statistics (NCES) through the Institutional Postsecondary Education Data System (IPEDS).  That data is made available to the public and to researchers.  The Department of Education reports an institution’s official graduate rate as the number of first-time, full-time freshmen students who graduate within 150 percent of the normal time to complete a degree program (three years for a two year degree and six years for a four year degree).  If an institution does not have any first-time, full-time freshmen, it will have a graduation rate of zero.  Many institutions that serve a large population of adult students will have very low numbers of first-time, full-time freshmen.  A useful source of data to determine the relative percentage of part-time students at an institution is to divide the FTE (full-time equivalent) students reported in a given year by the unduplicated headcount reported by the institution for the same year.  The lower the decimal calculated, the higher the number of part-time students will be who attend that institution.  If the majority of an institution’s students are part-time, the Department collects data on completions, but does not collect data about the length of time it took to complete a degree or the year in which the students who completed a degree matriculated at the institution.

Enrollments at colleges and universities that offer online programs have soared in recent years for many reasons, one of which is the convenience of working online from work or home versus commuting to a physical location for a class that may be held at times inconvenient for a working adult.  However, the consumer-friendly data that is collected by NCES relates to students who plan to be first-time, full-time freshmen and does not reflect the success rates of adult-students who matriculate and graduate from these programs.  Given that the department tracks unduplicated student headcounts and completions each year, adding two fields relating to unduplicated new students and drops for each year would provide a cohort-tracking system that could provide an aggregate completion/graduation rate for part-time students.  Many institutions serving part-time students allow seven years for those students to complete a two-year degree and up to ten years for a four-year degree.  The pace at which students complete those degrees varies based upon professional and family requirements that may conflict at times with the number of courses that a student can take.  One of the desired outcomes of our research was to provide a benchmark for similar studies as well as a suggested guideline for tracking persistence in adult-serving online degree programs.  As more and more adults return to college to complete their degrees online, finding a consistent system for measuring and reporting their progress will become more important.

Subjects of Interest


Higher Education

Independent Schools


Student Persistence