Introduction: Attrition is a Persistent Problem

Distance education has evolved greatly in the past decade, from paper based correspondence courses, to televised (or video taped) courses, to Computer Based Training, to web based courses, to interactive multimedia such as streaming video, each changing the way students participate in education. Whatever the format or distribution method, attrition has always been an issue. Attrition is also known as the drop out rate or non-completion rate, which can be defined as the number of students who had enrolled in a course but do not fulfill all the course requirements nor complete the course. The attrition issues in distance education suggest the following questions:

1. Why do students not complete the courses?
2. What can be done to decrease the attrition rate?

Distance education is not limited to K-12 and higher education; distance education has gained popularity in the corporate world as well. This is because distance education offers increased efficiency and reduced cost for both the student and their employers (Hall, 1997, Khan, 1997, and Zolkos, 1999).  An employer no longer has to pay transportation and lodging costs to send an employee to training, instead they can take an online course and receive similar instruction that they would in a face-to-face class. As the corporate world embraces distance education as a means to provide employee training, attrition rates are an important issue because training is perceived as not only a method to improve job satisfaction but also a method to improve the competitiveness of the corporation. If students do not complete the course, not only is valuable time lost but the student has lost the opportunity to acquire needed skills required by the corporation.

Many corporations are finding that there is no significant difference in learning performance whether in distance education or in face-to-face courses just as there is no significant difference in higher education institutions (Russell, 1998 and Zolkos, 1999).  One important difference between distance education in the corporate world and higher education is that training, whether distance education or face-to-face, is often a requirement to maintain or enhance job opportunities whereas in higher education, distance education is often, optional. An employee in a corporation must have the appropriate training to do their job, whereas a university student could opt to take another course or switch to a different college program. Therefore, in the corporate world, education is often mandatory for job performance and not completing a course can impact the employee’s job rating and future career opportunities.

Research: Goals

The goals of this research were to:

1. Determine the attrition rates for two distance education courses.
2. Compare the attrition rates of these courses with other corporate and higher education distance education programs.
3. Identify the reasons for attrition by collecting data from students.
4. Suggest methods based on analysis of the data to reduce student attrition.

 

Attrition Rates: Reasons

For as long as there have been distance education courses, researchers have tried to pinpoint reasons for student attrition. Researchers determined that the decision to complete a distance education course depends on a variety of interrelated factors such as age, marital status, educational level, and gender, which are particular to an individual’s context. (Morgan, 1999). Other factors that have been investigated as causes of attrition included number of courses and source of financial aid (Parker, 1999). Furthermore, the potential for a student to successfully complete a  distance education course also depends upon specific student characteristics that have been identified as indicators of potential success: 1) being a self-starter; 2) being self-disciplined; 3) being knowledgeable of the technology requirements of the specific format; 4) being able to meet other students and faculty in a virtual environment; and 5) wanting more control over learning environment. (Roblyer, 1999 and Wade, 1999). While these individual characteristics are important, it does not mean that there are not similarities in the reasons groups of students decide not to complete their courses.

Numerous studies have been conducted regarding attrition rates and reasons for attrition in distance education, but the majority of studies were conducted with data from higher education programs. This could be because corporations are less likely to openly acknowledge their attrition rates or otherwise divulge information that could benefit their competitors. Therefore the majority of the research cited in this paper is based on distance education courses conducted within higher education settings.

Distance education students are recognized to have a higher attrition rate than traditional students, (Thompson, 1997 and Phipps & Merisotis, 1999), although there has been much debate over the cause and how the attrition rates are calculated. Some universities do not consider a student who drops out during the official add/drop period as a part of their attrition rates, whereas other institutions consider any student who had enrolled, but not completed the course as part of their attrition rate. The Chronicle of Higher Education reported “No national statistics exist yet about how many students complete distance education programs or courses, but anecdotal evidence and studies by individual institutions suggest that course completion and program retention rates are generally lower in distance education courses than in their face-to-face counterparts” (Carr, 2000, p.A39). This suggests that ‘industry standards’ or accurate statistics on attrition are difficult to locate or non-existent.

A recent study on the difference between an Internet based Introduction to Computing Fundamentals course and a traditional correspondence course by the Open University in the United Kingdom conducted by Carswell et al. in 1998 found that attrition rates in the traditional correspondence course and the Internet based course were similar. The attrition rate in the Internet based course was 20% and the correspondence course was 16%, both of which were comparable to other Open University undergraduate courses. (Carswell et al., 2000, p36).  Carswell, et al.  stated that the distance format did not affect the overall attrition rate as the difference between the Internet and correspondence course was not significant.

Another study conducted in Illinois found that in their community college system, the Illinois Virtual Campus (IVC) network, some community colleges are reporting attrition rates greater than 20% when compared to their traditional classroom attrition rates. At Elgin Community College, 64% of the students completed distance education courses, compared to 83% who completed traditional classroom courses and Moraine Community College reported that for some distance education courses, the attrition rate reached up to 70% of the students (Breslin, 2000).

The American Institute for Chartered Property Casualty Underwriters (CPCU) and the Insurance Institute of America completed their first Internet based training classes in the spring of 1999 and reported that their results were similar to those of online learners in all subject matter areas. They reported that in some classes there was an attrition rate of up to 50%, but that the students who completed the class liked the online approach, learned the material, and spoke highly of the experience, despite the attrition rates. (Zolkos, 1999). 

This research suggests that higher attrition rates for distance education courses are common, but the majority of researchers do not provide conclusive reasons for the attrition rates.  Some of the reasons suggested were demographics, such as the student’s age and maturity level, but also time, and how experienced the instructor was with teaching online. (Carr, 2000). In general though, there is a lack of validated variables or frameworks to measure attrition within distance education courses (Sheets, 1992, Thompson, 1997 and Parker, 1999).

Phipps and Merisotis (2000, p. 21) identified 24 benchmarks needed to ensure a successful Internet- based, distance education course and identified one missing benchmark as attrition. The authors noted that in interviews with six higher education institutions, attrition was mentioned as an issue and the authors stated that “student attrition in Internet-based distance education courses is an important research topic in the evaluation and assessment…” of distance education courses.

 

Background Information: Advanced Skills Program

The Advanced Skills and Professional Development department was formed in November, 1998 for the purpose of increasing the skills of the IBM Printing System Division personnel who provide post-sales support for the solutions that IBM Printing System Division provides it’s customers. These solutions are combinations of physical hardware (printers and/or servers) and printing software (print management, formatting etc.) that are installed in the customer’s environment to provide complete print management.

 

Course Descriptions: Distance Education Offerings

The Applied Skills and Professional Development department has offered two distance education courses: Infoprint Manager Overview and Printing Fundamentals course to over 3,900 IBM employees worldwide using Lotus LearningSpace for development and deployment. Lotus LearningSpace is a complete and flexible e-learning platform that is able to deliver self-directed, asynchronous collaboration and/or  “virtual classroom learning” experience. LearningSpace can be delivered via the World Wide Web or Lotus Notes and is scalable, meaning LearningSpace can be located on a single server or in multi-server configurations, thus permitting either a single section or multiple sections of the same course to be taught at the same time.

These courses are delivered exclusively via Lotus Notes on the IBM Intranet for these reasons:

·      Ease of use and familiarity as students also use Lotus Notes for their e-mail. The course content is delivered as databases which the students are familiar with from using e-mail and other company applications, such as Human Resources.

·      Enhanced security as enrollment and access to course material is controlled by Lotus Notes security functions.

·      Availability to course material is 24 hours a day, 7 days a week, thus ensuring students can access the course material at anytime, in any time zone.

The students enrolled in the courses are required to download the courses onto their workstation or laptop via Lotus Notes and complete the course material and quizzes in a required amount of time. With successful completion of the courses, the student’s training records are updated and they are sent a certificate of completion.

Table 1 Course Descriptions

Course Description	Duration, Time to Complete, Enrollment, and Delivery Method
Printing Fundamentals: overview of printing industry and history of printing technology. Prerequisite for Infoprint Manager Overview. The courses includes online exercises, class discussions, reference materials, web links, and quizzes as part of the course materials. Students must successfully complete all quizzes.	Three weeks in duration and estimated time to complete is 24 hours. Enrollment: typically 30 to 60 students with approximately 30 to 50% of students from outside of North America. Delivered via Intranet using Lotus Learning Space and Lotus Notes
Infoprint Manager Overview: overview of Infoprint Manager software including functions and features. Prerequisite for instructor lead courses on Infoprint Manager. The courses includes online exercises, class discussions, reference materials, web links, and quizzes as part of the course materials. Students must successfully complete all quizzes and a final examination.	Two weeks in duration and estimated 16 hours to complete. Enrollment: typically 20 to 60 students with approximately 30 to 50% of students from outside of North America. Delivered via Intranet using Lotus Learning Space and Lotus Notes.

 

Courses: Attrition Rates

Attrition is defined for our purposes as: students who did not download the course, downloaded the course but did not complete any units, and students who failed to complete all quizzes and/or final examinations. The attrition rates for the two distance education courses are for a period of nine months:

Table 2 Course Attrition Rates

Course	Attrition Rate
Printing Fundamentals	23%
Infoprint Manager Overview	19%

While the overall attrition rate range, from 19 to 23%, is not as high other distance education courses cited in this paper and given that there are no existing standards for attrition rates in either corporate or higher education, the attrition rate was still considered an issue as these courses were prerequisites for other courses. Thus if a student did not complete these courses, they could not take a follow on course in the program curriculum.

Furthermore, given the program mission of improving employee skills, if employees do not complete the courses, then they have not received training that is essential to job performance, which directly effects customer satisfaction. Consequently, improvement in attrition rates is not only beneficial to the education program, the student, but also to the corporate goal of improved job performance, improved employee skills, and customer satisfaction.

 

Survey: Strategy

To gather student opinions and preferences, IBM commonly uses electronic surveys delivered via the World Wide Web or Lotus Notes with a high degree of success; therefore this format was chosen to distribute a survey to the employees who did not complete the courses. As stated earlier, students outside of North America who enrolled in the courses represented approximately 30 to 50 percent of all enrolled students and these students were from Asia, Europe, Central America, Latin America, and North America. Therefore, the survey would need to be sent to students, world-wide.

In selecting a survey tool and distribution method, the ability to contact students world wide was important and a common factor was that each student has access to Lotus Notes. For this reason, Lotus Notes was chosen as the tool to develop and distribute the survey. Paper surveys were ruled out as it would be expensive to mail the surveys world wide and would take much longer to obtain responses. With electronic surveys, responses are usually received within 24 hours.  Phone survey was ruled out because of expense and difficulty in reaching students in multiple time zones. Selective phone interviews were conducted to both validate the survey data sent via e-mail and to improve response rates as recommended by Salant and Dillman (1994).

 

Survey: Design

Possible topics for the survey questions were identified from a variety of sources: instructor feedback, student comments during the course, student end of course survey, and research in the field of distance education.

The student feedback came from an end of course student survey where students were asked a number of questions on the distance education course they had completed. All comments and feedback are then compiled into a course report, which is stored in a department database on Lotus Notes. From these sources of information, the possible areas of interest identified were download problems, time constraints, language skills, relevancy to their position, prior computer based learning experiences, and reason for taking the course. It is important to remember that these issues were concerns to students who had completed the course, whereas those who did not complete the courses did not have an opportunity to offer any input as they did not complete the end of course survey nor posted any comments during the course.

Following survey design and processes recommend by Salant & Dillman (1994), Preece, et al. (1994), and Shneiderman (1998), the survey questions were written using a Likert Scale to measure student opinions. The proposed survey questions were distributed to the Advanced Skills and Education department and after review comments were completed, an electronic survey was written using Lotus Notes Designer. The electronic survey was tested for ease of use and completeness by the Advanced Skills and Education department. After revising the electronic survey, the survey was distributed via Lotus Notes to the 62 students who had not completed a course. The survey was titled ‘Distance Education Survey’ and did not contain the words attrition to avoid offending any of the employees who did not complete the course by labeling them ‘drop-outs’. The complete survey is presented in Appendix A: Distance Education Survey.

In addition to the survey, telephone interviews were conducted by randomly selecting one in every six students including students located world-wide. The purpose of the telephone interview was twofold:

1. Validate student responses received via electronic survey. Another advantage of phone interviews was to provide students with a method to provide comments that they would not provide with the electronic survey.
2. Improve response rate.

The students were asked the same set of questions in the mail survey and if they had any additional comments that they would like to make. The results of the telephone interviews were disappointing.  The majority of the employees that had enrolled in the course were Customer Engineers and they generally do not work in an office, but rather on site at customer locations.  Therefore they are often difficult to reach via telephone. An e-mail was sent to those employees who had been randomly selected and who were off-site asking them to respond and this e-mail led to a few additional phone interviews.

In addition to phone interviews, an e-mail was sent two weeks after the survey was distributed to those students who had not responded asking them to respond. This e-mail provided a few more responses but in general, the phone interviews and follow-up e-mail did not elicit many responses.

 

Survey: Response Rate

Here are the survey response rates:

• A total of 62 students were surveyed and 35 responses were received for a response rate of 56%.
• Of the 35 responses, 8 responses or 23% were received from students who resided outside of North America.

Salant and Dillman (1994) recommended a minimum of 60 to 70% for any survey but to achieve this rate, Salant and Dillman stated repeated follow-up via e-mail and telephone is required. For this survey, given the lack of response to the telephone survey and follow-up e-mail, it was unlikely that any additional responses would be received.

Survey: Results

Question (See Appendix A for answer scale)	Results (N=35)
Q1. Were you able to successfully install the LearningSpace Courses?	Yes=80%; No=6%; Did not attempt=14%
Q2. Do you believe the time it took to install the LearningSpace course databases was acceptable?	Mean: 3.20 or 64%
Q3. Do you believe you had sufficient time to complete the course given your work schedule?	Mean: 2.44 or 49%
Q4. Do you believe your manager provided you with the necessary time to complete this course?	Mean: 3.11 or 62%
Q5. Did you work on the course at home, work, or both?	Home=15%; Work=27%; Both=58%
Q6. Do you believe the course took more time to complete than anticipated?	Mean: 2.97 or 59%
Q7. Do you believe the course content was relevant to your job?	Mean: 3.70 or 74%
Q8. How would you prefer to take this course?	LearningSpace=64%; Classroom=29%; WWW=0%; CBT=7%
Q9. Have you completed any other courses via distance learning such as using a Computer Based Training package, LearningSpace, or World Wide Web.	CBT=15; LearningSpace=13; WWW=8. (Number of total responses)
Q10. Do you believe the course was more difficult than anticipated?	Mean: 2.82 or 56%
Q11. Do you believe this course would be easier to complete if the content was in your native language?	Mean: 3.29 or 66%
Q12. Do you believe the reason you were enrolled in the course was clear?	Mean: 3.64 or 73%

 

Survey Results: Correlation and Descriptive Data

Data analysis was conducted using Microsoft Excel’s Data Tools Correlation function to determine if there was a linear relationship between the following variables:

• Q2=Instime (time to install the course materials)
• Q3=Comptime (sufficient time to complete the course)
• Q4=Mantime (manager provided time to complete the course)
• Q6=Time (took more time to complete course)
• Q7=Relv (course relevant to work)
• Q10=Diff (course more difficult to complete than anticipated)
• Q11=Lang (easier to complete if in native language)
• Q12=Reas (reason for enrollment was clear)

Correlation analysis was conducted on these variables and no significant relationships were shown to exist between any variables. Some variables (instime and comptime; comptime and mantime) indicated a positive relationship but the strength of the relationship was not significant enough to state any conclusions.

The following descriptive data was generated using Microsoft Excel’s Descriptive Statistical function:

Table 3 Descriptive Data

	Instime	Comptime	Mantime	Time	Relv	Diff	Lang	Reas
Mean	3.205882	2.441176	3.117647	2.970588	3.705882	2.823529	3.294118	3.647059
Median	3	2	3	3	4	3	3	4
Mode	4	2	3	3	4	3	3	4
Standard Deviation	1.148897	1.235612	0.977464	0.717119	0.718981	0.626224	0.759961	0.917254
Standard Error	0.197034	0.211906	0.167634	0.122985	0.123304	0.107397	0.130332	0.157308

 

Survey: Discussion

Before the survey was distributed, the following questions were postulated:

• Student dissatisfaction with the time to download would be very high based on comments students had provided in end of course surveys. It was assumed that this dissatisfaction would be a source of attrition for the courses.
• Students would want the course materials in their own language. Students where English is not their first language have commented that the course materials would be easier to read, given the technical nature of the material, if the material could be presented in their language. The students also believed that they could complete the course quicker if the material was in their language.

The survey data presented some interesting data points:

·      36% of students found the time to download the courses as unacceptable; this relatively low percentage was surprising as some students use modems (as opposed to higher speed connections available in offices and plant sites) to download the courses which can take almost three hours. It was anticipated this number would be much higher and would be a cause of attrition but 80 percent of the students had downloaded the course material successfully. Therefore, the time it took to download the course material was not a factor in student attrition though the time to download is a source of student dissatisfaction.

·      66% of students agreed that the course would be easier to complete if the course material was in their language but the data is misleading because the data, when broken out for the eight students who resided outside of North America, clearly indicated their preference for material in their language. This also matches student comments from the end-of-course surveys.

·      15% of students completed their coursework at home and 58% completed their coursework at both home and work. It was assumed that students completed the coursework at work not at their home. But since 51% of the students felt they did not have sufficient time to complete the course, it can  be assumed that their working on coursework at home may reflect a lack of time to complete coursework during work hours.

·      74% of students felt the course was relevant to their jobs; so the course content was not a factor in attrition rates.

·      Student comments from the survey and from the telephone interviews indicated that the primary reason for not completing the course was because of conflicts with work schedules.

Recommendations: Proposed Solutions to Reduce Attrition

Based on the data gathered, here are some proposed solutions to reduce the attrition rate:

• Provide additional links throughout the course content to a glossary to provide definitions of technical terminology to help all students, but especially students where English is not their first language. Additionally, we are investigating providing access to language translation dictionaries directly from the course.
• Extend the duration of the courses to provide students with more time to complete the courses; currently the courses are two and three weeks in duration. Extending the courses might enable students to spread the coursework out and may improve attrition rate but would have to be compared to the tradeoff of teaching a course longer which can impact the ability to update the course and frequency of scheduled courses.
• Promote the concept to both students and their managers of viewing the distance education courses as “virtual classrooms” and ask that students be given time to complete the coursework in the same manner as if they went to a traditional, instructor-led, classroom course.
• Permit students, on a student-by-student basis, to “temporary drop-out” out of one section of a course and then start another section of the course. Since we offer sections of the courses, almost every month, students can stop at one point and then start back at that point in another section. We do not encourage this as since we are promoting a virtual class, it is somewhat disruptive for a student to drop out of one section and then join another section but this does seem effective in reducing attrition.
• Reduce the file size of course material to help reduce the time it takes to download the courses onto student workstations or laptops. Even though a majority of students were able to download the courses and the time it takes to download the courses was not a significant factor in attrition, we believe that making it easier to download the courses might provide an incremental improvement in the attrition rate and would also improve overall student satisfaction.

Appendix A: Distance Education Survey

Were you able to successfully install the LearningSpace Courses?	Yes, No, Did Not Attempt
Do you believe the time it took to install the LearningSpace course databases was acceptable?	Strongly Agree, Agree, Neither, Disagree, Strongly Disagree
Do you believe you had sufficient time to complete the course given your works schedule?	Strongly Agree, Agree, Neither, Disagree, Strongly Disagree
Do you believe your manager provided you with the necessary time to complete this course?	Strongly Agree, Agree, Neither, Disagree, Strongly Disagree
Did you work on the course at home, work, or both?	Home, work, both
Do you believe the course took more time to complete than anticipated?	Strongly Agree, Agree, Neither, Disagree, Strongly Disagree
Do you believe the course content was relevant to your job?	Strongly Agree, Agree, Neither, Disagree, Strongly Disagree
How would you prefer to take this course?	Lotus LearningSpace, Classroom, World Wide Web, Computer Based Training
Have you completed any other courses via distance learning such as using a Computer Based Training package, a CD ROM, or from the World Wide Web.	Lotus LearningSpace, World Wide Web, Computer Based Training
Do you believe the course was more difficult that anticipated?	Strongly Agree, Agree, Neither, Disagree, Strongly Disagree
Do you believe this course would be easier to complete if the content was in your native language?	Strongly Agree, Agree, Neither, Disagree, Strongly Disagree
Do you believe the reason you were enrolled in the course was clear?	Strongly Agree, Agree, Neither, Disagree, Strongly Disagree
Comments:	Write in comments from students

 

References

Carr, S. (2000, February 11). As distance education comes of age, the challenge is keeping the students.  The Chronicle of Higher Education. pp. A39-41.  

Carswell, L., Thomas, P., Petre, M., Price, B., Richards, M. (2000). Distance education via the Internet: The student experience. British Journal of Educational Technology 31(1) pp.29-46.

Breslin-McSherry, M. (2000 June 11). On-line classes struggle to keep many students from logging off. The Chicago Tribune.

Hall, B. (1997), Web-Based Training, First Edition, Wiley Computer Publishing

Khan, B.H., (1997), Web-Based Instruction, First Edition, Educational Technology Publications

Morgan, C. K. and Tam, M. (1999). Unraveling the complexities of distance education student attrition. Distance Education, 20(1), pp.96-108.

Parker, A. (1999), A Study of Variables That Predict Dropout From Distance Education, International Journal of Educational Technology, V1(N2).

Phipps, R. and Merisotis, J., (1999), What’s the Difference? A Review of Contemporary Research on the Effectiveness of Distance Learning in Higher Education, Council for Higher Education Accreditation.

Phipps, R. and Merisotis, J., (2000), Quality On the Line, Benchmarks for Success in Internet-Based Distance Education, Institute for Higher Education Policy.

Preece, J., Sharp, H., Benyon, D., Holland, S., and Carey, T. (1994). Human-Computer Interaction, First Edition, England: Addison Wesley Longman

Roblyer, M.D., (1999), Is Choice an Important In Distance Learning: A Study of Student Motives for Taking Internet Based Courses at the High School and Community College Levels, International Society for Technology in Education, 32(1), pp. 157-171.

Russell, T. L.(1998) No Significant Difference: Phenomenon as Reported in 248 Research Reports, Summaries, and Papers Fourth Edition North Carolina State University Raleigh, North Carolina U.S.A.

Salant, P. and Dillman. D.A. (1994), How to Conduct Your Own Survey, First Edition, John Wiley and Sons.

Sheets, M. F. (1992), Characteristics of Adult Education Students and Factors Which Determine Course Completion: A Review, New Horizons in Adult Education, V6(N1), Spring,  1992., pp. 3-19.

Shneiderman, B. (1998), Designing the User Interface, Strategies for Effective Human-Computer Interaction, Third Edition, Addison-Wesley.

Thompson, E. (1997), Distance Education Drop-out: What Can We Do?, in Pospisil, R. and Willcoxson, L. (Eds), Learning Through Teaching, Proceedings of the 6^th Annual Teaching Forum, Murdoch University, February, 1997, pp.324-332. Available online: http://cleo.murdoch,edu.au/asu/pubs/tlf/tlf97/thom324.html. Verified 10/09/00.

Wade, W. (1999, October). What do students know and how do we know that they know it? Technological Horizons in Education Journal, 27(3), p94.

Zolkos, R. (1999, October). Online education getting good grades:  Despite high attrition, online courses seen as a possible alternative to the classroom.  Business Insurance, p. 40B.