Distance education drop-out: What can we do?

The extent to which a range of demographic, academic and administrative variables were related to attrition and persistence of external students enrolled in the Fourth Year of the Bachelor of Education course at Edith Cowan University (ECU) during second semester, 1995 were investigated in this study. Data were obtained from the student records system and two self-administered mail out questionnaires. Those students who withdrew from their studies had less teaching experience, had completed fewer units and semesters of study, and had lower course averages than the continuing students. The majority of students indicated that work, family and study commitments were the main reason(s) for their withdrawal. Withdrawn students were much less satisfied with the level of communication with the tutor and a greater proportion of these students rated assignment feedback as very unsatisfactory. The recommendations based on the results of the study are aimed at improving the effectiveness and efficiency of services offered to external students, and by doing so, reduce the attrition rate and provide educational benefits to those students continuing in their studies.

Introduction

The drop-out rates for distance education courses are usually higher than those for comparable on-campus courses (Kember, 1995). At ECU, the attrition rates over the last four years for external students have been more than double those for internal students. Furthermore, most student withdrawals occur in the early part of the semester and before the last day for withdrawal without financial penalty.

A study was conducted during semester 2, 1995 to investigate the extent to which any or all of the following variables related to attrition and persistence of external students enrolled in the Fourth Year of the Bachelor of Education course at ECU: age; gender; number of years of teaching experience; number of years since completing pre-service training; stage in the course (completed one or two of eight units); current occupation (teacher, principal, home duties, student); geographical location (metropolitan, country and interstate); method of communication with tutor, other students and external studies (electronic mail, post, phone, facsimile); administrative issues (quality of unit materials, late enrolment or receipt of unit materials, out of print texts, delays in assignment return); personal circumstances (e.g., separation from spouse); work related issues (e.g., change of school); perceived benefit of completing the course; and relevance of unit content to perceived career needs and interests. Data were obtained from the computerised student records system and two self-administered questionnaires.

Sample profile

There was no significant difference between the mean ages of continuing and withdrawn students, between these two groups in the mean number of years since the study participants completed their last teaching qualification or the number of previous withdrawals from units in this course (Table 1). Of the students in the study, 67.8% had not previously withdrawn from a unit, 15.5% and 9.3% of students had previously withdrawn from one and two units respectively.

There were significant differences between continuing and withdrawn students for the mean number of years of teaching experience (t = 2.43, df = 254, P = 0.016), the mean stage in the course (t = 3.21, df = 256, P = 0.002), the mean course average (t = 2.55, df = 256, P = 0.011) and the mean number of semesters successfully completed in the course (t = 2.22, df = 256, P = 0.027; Table 1). The withdrawn students had less teaching experience, had satisfactorily completed fewer units and semesters of study in this course, and had lower course averages than the continuing students.

There was a higher proportion of female students in the sample (84.5%), however, there was no significant difference between the genders for continuing and withdrawn students ( = 0.35, df = 1, P = 0.56). A third of the students (33.5%) would have preferred to have been enrolled in the internal or on-campus mode but there was no significant difference ( = 1.25, df = 1, P = 0.27) between continuing and withdrawn students in their choice of study mode. The majority of students were classroom teachers (66%) and there was no significant difference ( = 3.66, df = 5, P = 0.60) between continuing and withdrawn students for their current employment position. Most of the students lived in Western Australia (75.9%;) and there was no significant difference ( = 0.33, df = 2, P = 0.85) between continuing and withdrawn students for geographic location.

Fifty two (85%) of the 61 students in the study group who withdrew from at least one unit during the semester did so before the last date for withdrawal without financial penalty. A further eight students (13.1%) withdrew between week six and the end of week 10, the last date for withdrawal without academic penalty. Only one student (1.6%) withdrew after that time.

The students' written responses for reasons for their withdrawal were placed in six categories. "Work", "family" and "study" commitments were the most commonly reported reasons for withdrawal (Table 2).

Communication with tutor

Most students considered the tutor's comments about their assignments to be satisfactory or very satisfactory. However, there was a significant difference ( = 8.73, df = 3, P = 0.03) between continuing and withdrawn students for the rating of comments tutors made about their assignments. Fourteen percent of withdrawn students compared to 3% of continuing students rated assignment feedback as very unsatisfactory.

The results of this study confirm the views of the experienced staff in External Studies that the feeling of isolation for many distance education students is compounded when they are not informed who the tutor is for the unit that semester. The inclusion of an introductory letter from the tutor with the package of learning materials can be a reassurance for students, particularly for at-risk students (e.g., students enrolled in the early stage of the course or the less experienced teachers amongst them). The following comments are illustrative of this situation. "I was not assigned a tutor and needed help with the first assignment. I phoned the coordinator of External Studies twice but got no reply. I failed my first assignment", and "ECU did not have tutor until very late. As a result second assignment sent in before I found out who tutor was and before I received 1st assignment back".

Previous research (Sweet, 1986; Kember, 1989) indicates that direct telephone contact between the academic staff and the students has a positive influence on student commitment and persistence. With a high proportion of students making use of the telephone to communicate with the tutor, the effective use of voicemail by university staff is essential. For example, tutors can modify the voicemail message to address common concerns of students that arise during the semester.

Communication with the tutor is often made more difficult for external students due to their geographical location and the resulting time zone differences which, in Australia, may be up to three hours. Communication between the students in this study and the tutors may also have been hindered as many of them were classroom teachers and therefore not readily able to be contacted by telephone during school hours. Both these factors again support the use of voicemail facilities and probably, in the future, enhance the possible use of e-mail.

Relationship of variables to student attrition and persistence

Table 3: Discriminant analysis to determine the variables known at the commencement of the unit that are associated with withdrawal
Variables	SCDFC*	CORR**
Age	0.109	0.134
Gender	-0.121	-0.122
Geographic location	-0.177	-0.072
Stage in course	0.823	0.667
Course average	0.386	0.574
No of semesters satisfactorily completed	-0.282	0.490
No of previous withdrawals	-0.282	-0.114
No of years since completion of last teaching qualification	0.084	0.274
No of years teaching experience	0.420	0.462
Chosen mode of study	0.082	0.180
Perceived relevancy of unit materials	-0.107	-0.103
Level of satisfaction with external studies administration	-0.294	-0.271
* Standardised canonical discriminant function coefficients ** Pooled within-groups correlations between discriminating variables and canonical discriminant functions 69.35% of students correctively classified, Eigenvalue = 0.105, Wilks' Lambda score = 0.904, = 24.06, df = 12, P < 0.001.

A discriminant analysis using the variables listed in Table 4 correctly classified 74.44% of students according to whether they continued or withdrew from their studies. The variable "level of satisfaction with communication with tutor" was the factor identified in a stepwise discriminant analysis to account for most of the separation between the two groups and correctly classified 72.43% of the total students into whether they withdrew or continued with their studies. The variable "level of satisfaction with communication with tutor" correlated astonishingly highly (r = 0.99) with the canonical discriminant function that separated continuing and withdrawn students. The two most commonly raised criticisms by students were that they were expected to submit their next assignment before receiving feedback on the last one and, secondly, the comments on their assignments were unconstructive, negative and demeaning. Some examples of students' responses concerning tutors' comments about their assignments were "often lacked supporting feedback for mark allocation"; "tutors are always negative in their comments and concentrate on academic grades rather than understanding. They tell you what not to do but offer no alternatives."; and, "they were brief and highly critical and demeaning. They really made me feel as though I was incapable. No constructive criticism at all".

Table 4: Discriminant analysis to determine the variables known at the completion of the unit that are associated with withdrawal
Variables	SCDFC*	CORR**
Relevancy of unit materials	-0.062	0.26
Level of satisfaction with communication with tutor	1.005	0.99
Level of satisfaction with assignment feedback from tutor	-0.007	0.57
Level of satisfaction with communication with External Studies for administration issues	0.115	0.19
* Standardised canonical discriminant function coefficients ** Pooled within-groups correlations between discriminating variables and canonical discriminant functions 74.44% of students correctively classified, Eigenvalue = 0.023, Wilks' Lambda score = 0.977, = 5.03, df = 4, P = 0.28.

Relevance of student profile information

The results of this study suggest that the most at-risk students were those commencing the Fourth Year of the Bachelor of Education course in their first or second year of teaching. A number of student comments supported this finding. For example, "I started teaching for the first time this year after completing a Bachelor of Arts (Education) last year. The demands of the job made it impossible for me to find the time required to complete the course successfully". In addition, the results suggest that the more units a student had completed the less likely that student was to withdraw. These results were not unexpected as those students less suited to external study were more likely to withdraw during the early stages of the course, and those students persisting were expected to have an increasing commitment to completing the award with the satisfactory completion of each unit.

The variable "course average" was also strongly, positively correlated with the canonical discriminant function that separated continuing and withdrawn students and, as expected, withdrawn students had a lower course average than continuing students. Kember, Lai, Murphy, Siaw & Yuen (1994) suggest that student results (GPA) functions as an intervening variable, with some students who receive low grades being discouraged from continuing with their studies.

External Studies personnel and those staff teaching external units need to be advised that the more at-risk students in the Fourth Year of the Bachelor of Education course are those in the earlier stages of the award, who are achieving comparatively lower grades and who have not been working in schools for as long as other students. Particular attention by the university to the needs and problems of these students during the early part of the semester may reduce the number of student withdrawals. Knowing a student is in the "at-risk" category should not necessarily exclude them from a place in the course as there are a multiplicity of factors influencing a student's decision to withdraw (Kember, 1995). However, with the appropriate advice at the admissions stage and some modification of university procedures, these at-risk students may overcome adverse characteristics or circumstances. For example, tutor-initiated contact with students early in the semester (e.g., an introductory letter or phone call), may make all the difference to some students who are feeling unsure about how to tackle an assignment, whether the work load is too heavy, or where to locate resources.

Reasons for withdrawal

The clear identification of reasons for withdrawal is an acknowledged problem in attrition research (Price, Harte & Cole, 1992). Furthermore, students are often reluctant to identify the real reason(s) for withdrawal, there may be multiple reasons for them discontinuing or the reason(s) given may not reflect the underlying difficulties a student is encountering (e.g., "returned to teaching" may be due to increased financial pressure). Some caution is therefore appropriate when considering responses to questions relating to "reasons for withdrawal" in this and other attrition studies.

Many students suggested there were multiple reasons for their withdrawal, however, the majority of students indicated that work, family or study commitments were the main reasons for the withdrawal. Two typical student written responses to this question were "I just do not have the necessary hours in the day to teach full-time, run a house, three children, a husband and a dog as well as study and remain sane" and "family life and commitments were drastically affected - kids thought Mummy was always doing assignments - I was unable to do the things I wanted with and for the family". Another student explained the decision to withdraw with the comment "increased demands at work, coupled with personal dramas made it difficult to allocate adequate time and concentration to study".

A number of students' comments in response to the question relating to the reason(s) for withdrawal indicated that administration issues, such as the late receipt of unit materials and the lack of feedback from the tutor on their first assignment by the Higher Education Contribution Scheme (HECS) assessment date, contributed to their decision to withdraw. For example, "withdrawal mainly due to lack of organisation in personal life but compounded by unit material not forwarded until two weeks after semester commenced" and "if I had received my first failed assignment back before the HECS date I would have withdrawn before that date, therefore no charge".

Perceptions and difficulties of studying externally

Seventy five percent of withdrawn students indicated that "work" had an adverse affect on academic progress. There may be a need for the university administration to be more flexible in the structuring of courses and the timing of units offered. For example, as most of the students in the study were teachers, university examinations conflicted with the writing of school reports and other major administration tasks required to be completed at the end of the school semester. Units in the Fourth Year of the Bachelor of Education award may, therefore, be more appropriately offered during school vacation periods or there may be a need for more continuous assessment in the course.

Summary

References

Billings, D.M. (1988). A conceptual model of correspondence course completion. The American Journal of Distance Education, 2(2), 23-35.

Cookson, P.S. (1989). Research on learners and learning in distance education: A review. The American Journal of Distance Education, 3(2), 22-34.

Garrison, D.R. (1987). Researching dropout in distance education. Distance Education, 8(1), 95-101.

Kember, D. (1989). A longitudinal-process model of drop-out from distance education. Journal of Higher Education, 60(3), 278-301.

Kember, D. (1995). Open Learning Courses for Adults: A Model of Student Progress. Englewood Cliffs, New Jersey: Educational Technology Publications.

Kember, D., Lai, T., Murphy, D., Siaw, I. & Yuen, K. S. (1994). Student progress in distance education courses: A replication study. Adult Education Quarterly, 45(1), 286-301.

Price, D., Harte, J., & Cole, M. (1992). Student Progression in Higher Education: A Study of Attrition at Northern Territory University. Canberra: Australian Government Printing Service.

Sweet, R. (1986). Student drop-out in distance education: An application of Tinto's model. Distance Education, 7(2), 201-213.

Woodley, A. & Parlett, M. (1983). Student drop-out. Teaching at a Distance, 24, 2-23.

Zajkowski, M.E. (1992). Recent research at The Open Polytechnic of New Zealand (TOPNZ). Research in Distance Education, 4(4), 5-6.