Official and Unofficial Withdraw: Student Perceptions and Trends for Withdrawing from a Course.

Hyllegard & Deng (2008) set out to examine the reasons why attrition rates from online courses were high at the Borough of Manhattan Community College in New York.  To accomplish this they collected information from the online students in the fall of 2003.  Of the original 685 students, 190 withdrew from withdrew from at least one online course.  The overall attrition rate for the online sections of these courses was 26% compared to 12% in traditional sections of the same courses. Hyllegard & Deng received survey responses from 56 of these students; 35 who had officially withdrawn and 21 who had unofficially withdrawn.  Students who unofficially withdraw did not complete the college’s process for withdrawing from a course, resulting in a failure of the course or receiving an incomplete for the course.  Both Students who officially and unofficially withdrew from courses cited lack of time or personal problems (61%), the online format of the class (30%) and technological issues (9%).  They expressed that while they had enough time each week to keep up with their courses that the classes had required more reading and writing than they had expected.

When compared to the college student body as a whole, students who officially withdrew were more likely to have registered for a heavier course load, have a higher GPA, be a full-time student, and to have completed their developmental course requirements.  They determined through examining course logs, that students who had officially withdrawn had logged on to the course 10 or more time, which indicated that they made a concerted effort to engage with the course materials before choosing to withdraw from the course.  Hyllegard & Deng conclude that students who officially withdrew were more likely to have been overextended and that demands in their personal life required them to reduce their temporarily reduce their course load in order to.

In contrast, students who unofficially withdrew were more likely than those who officially withdrew to be Hispanic, have a lower entering GPA, enroll in few credits, to have not completed their developmental coursework, and were more likely to take a higher proportion of their courses online.  These students were less engaged with the course material, and were more likely to report withdrawing due to the online format (40%).  These students also had far more dismal outcomes for the semester with an average semester GPA of only 0.6 and having on average a 0.23 ratio of completed to attempted course credits.

Interestingly, 60% of both student groups reported that they would enroll in an online course at Borough of Manhattan Community College again.  However, as the authors point out, those students who did not officially withdraw will be at a distinct disadvantage due to the impact their failing grades have on their overall academic record.  The authors conclude that those students who did not officially withdraw, in contrast with those students who officially withdrew, are more likely to be at-risk students who struggling with their course work had “hoped online education would be easier, less demanding route to academic success.”

GPA, Computer Skills, and Auditory Learning Styles: Student Characteristics Linked to Failed Persistence in Online Courses.

Harrell II and Bower (2010) studied the characteristics of students that predict persistence in community college online courses.  They identified a trio of three characteristics that predicted withdrawal or failure in online community college courses.  They utilized an online survey instrument that included the Barsch Learning Style Inventory, the Abbreviated Measure of Internal/External Locus of Control, and a computer experience scale.  The researchers utilized confirmatory factor analysis and exploratory factor analysis to support the validity of their instrument and Cronbach’s alpha coefficients (of each section) to determine support the reliability of their instrument.  This survey was distributed to a total population of 1,593 students in five community colleges in Florida.  Participation in the survey was voluntary.  544 survey responses were received, 225 were considered valid cases having met the criterion of the study (complete and having enrolled in an online course the semester previous).  The survey captured the following student characteristics: 

• Demographics: gender, GPA (at the time the online course initiated), age, race, enrollment status, marital status, employment status, number of children, financial aid status
• Learning Style (Barsch):  Auditory, Kinesthetic, Tactile, Visual
• Internal/External Locus of Control
• Experience with computers
• Experience with online coursework

The researchers defined persistence as “whether or not a student completed the online course(s) in which he or she was enrolled during the previous semester.”  The researchers do not indicate if this information was provided by the institutions or by the students as part of the survey.

Harrell II and Bower utilized Stepwise Logistical Regression, ANOVA, and t-test analysis to isolate the characteristics that corresponded with students who did not persist in their online educational course(s) in the previous semester.  They tested both one-variable and three-variable models for fit, and determined that a three-variable model had the best fit.  Utilizing logistical regression analysis they determined that GPA, Auditory learning styles and computer skills were significant characteristics for predicting student persistence in online courses in community colleges.  Auditory learning style, and computer skills are negative predictors of online course persistence and GPA was a positive predictor of online course persistence.

As a result of their findings they recommend that community colleges identify students with who have a low entering GPA and are lacking basic computer skill as at risk for withdrawal (or failed persistence).  They suggest that community college provide supportive resources for those with a low GPA “to develop the academic skills necessary to perform well in their courses” and that these students be enrolled in a “student success course” preparing them for online courses.  They also suggest that community colleges increase the utilization of audio files, live chat sessions that accommodate audio, and streaming video to help promote learning for those students with an auditory learning style.  The researchers make no recommendations regarding addressing students’ who lack the basic computer skills necessary to be successful in the online courses, although it might be assumed that these needs would be addressed in the “student success course” recommended for students with a low GPA.

Harrell and Bower do an excellent job of exploring the available research on online education in community colleges; however, they do not explain the significance of the problem in this setting.  Why did they chose the community college setting, why persistence, and why in online courses?  They also do not describe what online coursework the students in this study were engaged in.  Was it entry level coursework, advanced, asynchronous, synchronous, credit or non-credit, open access, selective, etc. 

Unfortunately, the authors do not provide the information a practitioner would need to implement their recommendations.  For example, how do you know how judge if your program needs more auditory sensory information? How much auditory sensory information was used in the courses studied?  Would a “student success course” work for students attempting introductory coursework only or would it be applicable to complex or advanced coursework settings?  How do we know that a “student success course” intended to help them be successful in online coursework is the right focus? How do practitioners determine what the GPA cut off should be?  Would it be different for different programs? Might students with auditory learning, low GPAs, and poor computer skills be at risk in all of their classes?  Do we know if these students withdrew from predominately their online coursework?  Overall, Harrell and Bower research methods are solid and they provide a good foundation for further inquiry by both practitioners and researchers.

Factors Contributing to Online Course Completion

Aragon and Johnson (2008) studied factors that influence completion of community college online courses.  They identified three factors that contribute to persistence in courses.  Students who are female, enrolled in a higher number of online courses, and/or who are entered the course with a higher GPA were more likely to complete online courses.  Additionally, they explored the reasons that students withdraw from online courses.  Students reported withdrawing from courses because of: personal reasons (34%), course design (28%), technological reasons (18%), institutional mistakes (11%), and learning preferences (9%).

Data was collected from 305 students taking an online course at a rural community college in the Midwest.  The researchers utilized a combination of data that was collected as part of the organizations data management.  These included: student demographic, enrollment, academic readiness, and course completion.  The researchers defined completion as receiving a grade of A, B, C, or D, and academic readiness as a combination of GPA and academic placement test results.  Measures of self-directed learning were measured utilizing Bartlett-Kotrlik Inventory of Self-Learning, which was administered online.  They also contacted via telephone 65 of the 116 who withdraw from or failed their online course.  These students were asked “Why did you choose not to complete your online course?” (p. 149). 

The researchers utilized a chi-test to compare student demographics and academic preparedness between completers and non-completers, and a t-test to compare age, hours enrolled, hours enrolled in online courses, GPA, self-directed learning variables.  They subsequently utilized the Pearson product-moment correlation coefficients to determine the strength and direction of the relationships among the variables.

As a result of their findings they recommend that community colleges advisers identify students with who have a low entering GPA and provide them with additional counseling on the workload and time management of online courses.   Unfortunately, the authors do not provide practitioners with a means to determine what the cut off should be for a low entering GPA.  Aragon and Johnson recommend utilizing a “student help desk” to assist students in addressing technology related issues including provide computer related trainings as needed.   The authors do not address if this help desk should be online or a physical space on the campus, and if utilization of this service should be optional or mandatory for students. Additionally, they recommend that “instructors need to set the parameters of when communication is to occur and then abide by those times” (p.155). 

Citing that a significant portion of the sample of students who withdrew did so because of course design (28%) the authors recommend that course designers need to be cognizant of the unique demands of the online educational environment and its students.  They suggested that an innovative approach to course design that recognizing differences among the learners, motivates students to learn, avoids overloading the student, is based in a real-life context, and that provides both hands-on activities and opportunities for student reflection (p.155).   Which leads to their final recommendation which is for community colleges to establish ongoing measures for quality control and for oversight of course development.

Community Colleges: Market Leaders in Online Education

 

The above video illustrates the prevalence of online education in community colleges in the United States.  The fact is that community colleges are the leading provider of online education, and that this prevalence is all the more striking when you focus on the undergraduate course offerings (Allen and Seamus, 2008; ITC, 2010; NCES, 2008).  Additionally, despite continued growth, student demand for online education has continuously exceeded availability (ITC, 2010).  The growth of the online education industry is outpacing the research needed to build efficient and effective online student programs, especially in community colleges. 

It is important that we do not assume that the tools designed for and research done with traditional students on four year campus will be effective when used with the online community college student. Online students, community college students, and online community college students are all more likely to be what is termed “nontraditional.”  Online students are more likely to be 23 years or older, married with children, employed (often full time), and racially diverse (Reisetter and Boris, 2004; Wyatt, 2005).  In addition to all these characteristics, community college student are also more likely to be less academically prepared and come from lower income families than undergraduates in four year institutions (Cohen, 1990).  Community colleges report that 50 percent of their online students are twenty-six years of age or older, and that sixty three percent are females (ITC, 2010).  The demographics described in these studies suggest that current instruments, methods, and findings from studies utilizing traditional students, may not meet adequately or accurately represent the majority of online students and the programs that serve them.

As a doctoral student, I am pursuing this line of research, and I know that I am not alone.  I am hope though this blog to highlight some of the amazing work being done, connect some of the people doing it, and gain insight from you.  While my plan is to focus on highlights of relevant research, I will also provide relevant links and occasionally share some thoughts on online technology trends, etc. 

I think it, is imperative that researchers recognize and respond to the unique needs of community colleges virtual and online programs as well as the students they serve.  It is also important that this research is timely and distributed to the administration, faculty, and staff providing online coursework to students.  I also want to encourage anyone currently working in this field to suggest areas for research, and for anyone who is doing research in this area to contact me so I can highlight your work.

Research, the CC Online Blog, and the Nine-tailed Fox

Foxes appear in the folklore (both past and present) of nearly every culture that has come into contact with them.  The roles of these mythical foxes are often both nebulous and multidimensional.  Among the many roles the fox has been cast are the loyal protector, the intelligent strategist, the elusive seductress, the judge and jury, the murderous monster, the shape-shifter, and the mischievous thief.  It is not unusual to find the fox as both the object of great fear and admiration.  The origin of the nine tail fox is of some controversy.  The nine-tailed fox appears in mythologies originating from many Asian countries including China, Korea, India, and Japan. Perhaps the best known are the Japanese fox spirits or Kitsune.  Living in some cases for hundreds of years, they grow up to nine tails as they age and become wiser and more powerful.  These mythical foxes are depicted as benevolent, mischievous, or even malicious.  

Similarly, research is perhaps one of the most nebulous concepts of our time.  Research is often honored as respectable and creditable, but is also met with skepticism and distain.  Viewed by some as a product of elitism, research can be viewed as out of touch and inflicted upon people.  However, research is often considered the primary product of intellectual pursuit and is cherished by many as the source of modern knowledge and innovation.  On one hand, research can provide us a statistical description that is both accurate and that fails to represent any individual member of the population in question.  On the other, research can examine a small population so intensely that it is difficult to know the potential relevance of this knowledge beyond the immediate population that was studied. 

The range of quality in research, the research intent and biases, and the biases of those who distribute and utilize the results of research, create a setting where research is not value neutral.  

The ethical dilemmas that surround research are complex, and are often based on very subjective interpretation of the associated risks and benefits.  This, in combination with past ethical transgressions, has led some people to recognize that there is a dark side to research that evokes fear and mistrust.  I would argue that a large portion of this mistrust results from how research is distributed outside of the research institutions that create them. Often, practitioners and the public at large do not have access to the publications that result from research.  Even when they do, the language, depth, and methodologies utilized render the research inaccessible by the average reader.  This is not to insinuate that the average reader is not intelligent enough to understand research, only that they rarely have the resources (time, motivation, and specialized knowledge) to do so.  Instead they rely on other mediators, who unfortunately often carry a heavily bias, to select and interrupt research for them.  The quality of this process varies greatly and in some cases it can be more reflective of the motivation of the mediator then the research.  Just as importantly, there are huge bodies of quality research that are never distributed outside of the research institutions that conduct them. 

Researchers struggle with limited access to practitioners and a sometimes outdated or an abstract understanding of the relevant problems.  They too are influenced by available resources (access to sites, funding for their studies, and time).  While some researchers may be arrogant or reckless, I believe most are daunted by the potential impact of their research.  This apprehension leads them to be cautious about sharing their work outside of the research and academic institutions where they are comfortable.  There does appear a very real desire among researchers for their research to make an impact outside of the research institutions they are serving.  However, like their counterparts (the practitioners) researchers too often lack the resources needed to distribute their findings (time, funding, knowledge and accesses to mechanisms, etc.).  Additionally, rarely are reward mechanisms supportive of the efforts of either practitioners or researchers to bridge this divide.

My intent with this blog is not to change the nature of the fox.  Instead, I am hoping that those of us audacious enough to face the nebulous nature of research can build a network that will provide us the tools and collaborations necessary to be effective stewards of our work.  The purpose of this blog is to build a space where researchers, practitioners, and those impacted by online learning in community colleges can learn about research that is relevant to their work, share ideas for future research, and where we can network and learn from each other.