Computers & Geosciences, Volume 25, Number 10, 1999

John W. F Waldron
Geology Department
Saint Mary's University
Halifax Nova Scotia Canada B3H 3C3

John C. Butler
University of Houston
Houston, TX 77204

From the Associate Editor

I was impresssed with John Waldron's web pages the first time I found them. I knew that he had taught an online course and asked that he share some of his observations with the readers. I know that there is a lot of talk and some action around the globe about online courses in general and about geosciences courses in particular. The conclusion of his editorial will appear in the next issue.

Experiences teaching an introductory geology course on the web..Part II

Initially, we intended to use WebCT's quiz module to gather the results of assignments. For practical reasons I more or less abandoned this approach as the course progressed, and had the students send their answers to me by email. We reserved the WebCT testing tools for two purposes. Short, automatically marked 'self-tests' were provided for many individual pages, to help the students test their understanding but with no marks recorded. Longer multiple-choice and short-answer tests were delivered by computer as part of a mid-term test and a final test for each semester of the course, and we would arrange a proctored situation for these more formal exams.

WebCT provides a plethora of other tools, and we presented a selection for use by the students. Most important were the electronic mail and bulletins. Both these tools had the same interface, the only difference being that bulletins were automatically sent to all members of the class. Most of the bulletins that were presented came from me - they were general announcements about due dates, changes to the schedule, and clarifications where an individual student's enquiry showed a need for some additional information to be broadcast to the whole group. We had a chat tool which was a little primitive. During testing of the course I found our volunteer testers to be so enthusiastic that the chat discussion was almost impossible to manage: it was impossible to answer one question before another came in. I therefore held off from enabling the chat tool until the course was well underway. There is an excellent glossary tool, that enables the instructor to hyperlink technical terms in the on-line classes with a click of a button. We were fortunate to receive permission from the University of Iowa to use its web glossary of geology as the basis for the development of ours.

When the big day came for the start of classes, I found about 20 students registered - not a large number compared with the numbers in the classroom based courses. I say 'about 20 because there were several late registrants and people who completed some parts of the registration process but not others, so it was several weeks before I knew exactly which students would become permanent residents on the course pages.

It rapidly emerged that the students were not, in general, members of the group we had originally envisaged taking the course: distant students or those thinking of re-entering full-time education. Instead, most were already full-time students resident in Halifax, and seeking to fit an additional course into a full timetable of classes. In other words, they matched closely the population that typically filled the regular lecture classes during the semesters. Certainly this had some advantages: in some cases where technological communication failed, we could fall back on the telephone or a face to face office visit to fix up a problem. All the students were able to come to the Geology Department to pick up their sample kit, and all were able to come to the four formal tests without our having to worry about arranging proctored tests at remote locations. But there were disadvantages as well: some of the students wanted to get a quick credit without too much work, and they probably looked jealously at their peers in the conventional classrooms, for whom the assignment load was lighter.

Because all the students were close together, sometimes the degree of collaboration on assignments crossed the line that separates useful help from plagiarism. Under those circumstances I had to issue warnings that all work must be in the students' own words. I took the same approach I would have used in the classroom. For an initial occurrence, especially where the problem was overly close collaboration on wording rather than exact copying, I would issue a private, and friendly, warning pointing out the text that I though was too collaborative, and indicating where I placed the boundary between 'helping each other learn' and 'copying'. My first such warning was met with a vehement and strongly worded denial from the students involved, together with a claim that my allusion to the apparently copied text (both versions contained the same mis-spelling of a word) was condescending and inappropriate. Both I and the students learned very fast that words on a computer terminal can look very threatening compared with the same words conveyed face to face, where they can be accompanied by appropriate gestures and non-verbal cues.

Meanwhile, my colleague in the Astronomy course, using a series of multiple-choice tests, had already caught several students taking the test while they had pages of course notes simultaneously open on their computer. One student was dismissed from that course as a result of repeat offences, and these problems were eventually solved to my colleague's satisfaction; nonetheless, the possibility of systematic cheating confirmed my preference for a more assignment-based evaluation of the students, with a small number of supervised exams.

Student responses to the course were quite varied. Early in the course, students tended to give feedback only where they found problems, either with the technology or with the content. As a result, most of the feedback I received was negative, but I don't think this indicated any general level of dissatisfaction. In general, I found that I needed to do much more editing of my materials than in a normal class. In a classroom, if I make a mistake in a question sheet or on the overhead projector, it is relatively easy to congratulate the student who spots the mistake, announce the change, and move on. On a web page, mistakes seem much more embarrassing, especially if the students sit looking at them for a week before some brave soul sends in an email questioning the information! So I spent a lot of time correcting and finishing off the pages.

Assignments generated most of the feedback. In a number of cases in our initial design, an assignment was presented on the last page of notes for a module, but students then had to submit their answers on a form provided using the quiz tool within WebCT. Despite our efforts to customize this tool to say 'assignment' rather than 'quiz', we were unable to remove all of the more formal, examination-related 'look and feel' of these pages. One result was that the students found them intimidating. Also, in some cases, there were slight inconsistencies in the wording of questions or answers between the presentation on the regular pages and the more terse presentation of the questions in the quiz tool. These took the students by surprise: some felt they were being asked to do two different assignments. The final straw for the quiz tool was that the marking of longer answers proved to be extremely cumbersome in the version of WebCT we were using. (This has been much improved in the current version of WebCT). As a result of all these problems, after about six weeks I abandoned the quiz tool for everything except the simple multiple-choice and short answer questions for which it was designed, and had the students send their assignment answers to me by e-mail.

>From this point on, the assignments went reasonably smoothly. In fact, a proportion of the students surprised me with the enthusiasm and industry that they put into the assignments. A high point for me was the assignment that I gave for the students to pass in when they returned after their vacation in January. In the aftermath of hurricane Mitch and the eruption of Popo, I asked them to collect information from the world-wide web describing the geologic hazards affecting Central America, incorporating both tectonic hazards and those from the oceans and atmosphere. Some students incorporated graphical information from the web directly into their assignments - this took a little trial and error before we arrived at the best method of submitting the attached documents. Although some of the assignments were the usual minimal efforts, others were very well done indeed. One, in particular, I recall as being visually the most exciting assignment I have received from any student in any course I have taught.

I gradually learned more about the way in which the students were using the materials we provided. A number of students, it turned out, did not have computers in their residence rooms, but were trying to take the course using general-access computer labs on the university network. This approach was not specifically excluded by the requirements for the course - all we had said was that students must 'have access' to a computer of a certain specification, not that they actually had to live with one! I only became fully aware of this situation in December, when the university announced a long-awaited overhauled of billing procedures, with the result that laser printing (previously almost free) would be charged at 10 a page. A student asked me to put up all the remaining modules before December 31, because she would be printing them out systematically and taking them away for study! This would suddenly become very expensive on January 1 with the new billing procedures. Of course, this meant our laboriously produced animations and self-test questions were not being used by this student as she read the pages!

It also appeared to me from the answers to assignments and tests, and from a look at the times when students accessed each page (all exhaustively logged and available through WebCT), that some students would read the pages very rapidly, 'gliding over' the content without taking much in. This is a problem that I have plans to address in this years' offering, using some new facilities in WebCT to delay release of pages until certain pieces of work have been completed.

My fears about the chat room proved unfounded. When I finally enabled the WebCT chat tool, almost nobody used it, preferring to send in their questions by more private electronic mail.

A number of students dropped out of the course during the year. Either I would stop receiving the regular assignments or a student would fail to show up at one of the four tests. Most of these students eventually submitted withdrawal forms, but others simply failed to complete the course with the result that they received automatic F grades. The proportion was somewhat larger than the dropout rate I would typically see in a classroom-based course, but the majority of the dropouts occurred early on, and were clearly people who had practical difficulties with the web format. My understanding from accounts of web courses in other subjects (e.g. MacEachern 1999) is that our student retention was comparatively good. In the end 13 students appeared at the final test.

The students were required to hand in the rock sample kits at the time of the final exam, but we had given little consideration to the recovery of the kits from students who had dropped out. With the registrar's assistance, we were able to telephone those students and indicate that they would receive a bill for the replacement cost; as a result we were able to recover all the kits but one.

Overall, the grades obtained by those completing the course were similar to those I would expect in the classroom-based sections, but precise comparisons are not easy to make because of the very different proportions of marks assigned to assignments and tests.

Some aspects of the organization of the university affect the success of web-based teaching. First, support from information technology professionals is essential. In our case, the computer services department received a new name and underwent several personnel changes during the development and first delivery of our course, with the result that the support we received was erratic. At times, we would have excellent communication from 'ITSS'; then the person we were dealing with would move on, and the support would evaporate for a while. I would suggest that stability of the human knowledge-base on course delivery software is a major factor in determining the viability of a web-based course offering.

A second observation is that we really needed an assistant who was conversant both with web-page construction (including graphics) and with geology at a senior undergraduate level. Because we were unable to find such a person within the short time-lines imposed by the Notemakers program, we ended up with assistants who knew little of the subject material. This led to a repetitive cycle of correction (especially for the graphics) that made the project unnecessarily time-consuming.

A third consideration is uppermost in my mind as I face becoming chair of the department this year. The web course is expensive to offer and has a detrimental effect on our enrollment figures. To teach the course well, even to a small class, requires at least as much time input as a regular course, and counts as a regular course in the assessment of my teaching load. Currently, it looks as if this year's enrollment will be comparable to last year's. If the course were to expand, it would inevitably have to move away from assignments to multiple-choice automated testing with all its disadvantages. A hard-nosed economic climate prevails in the university, in which decisions on faculty replacements and other resources seem to be based heavily on total course enrollment. In these circumstances, though I believe we have developed an excellent and innovative course, there is a strong possibility that we will not be able to continue to offer it after the 1999-2000 year, unless there is a change in the way the university breaks down enrollment figures by department.

Overall, teaching on the web has been a valuable experience, and we will be offering the course again by the time you read this in the 1999-2000 year. We have made some changes based on our experiences last year. There will be a series of quizzes at the ends of modules. Though these will not count toward the final mark, the students will have to complete and pass each quiz before being allowed to progress to the next module. My hope is that this will force the students to pace their learning a little better. The chat is gone, and there are some 'read me first' pages on how to take a web course, giving advice, and information on what constitutes plagiarism. In the coming year I hope we will be addressing some of the institutional questions that are forcing us to look skeptically at the enrollment figures. And we have what we think is one of the best web sites in the world on introductory Earth science, but because of copyright limitations on the diagrams we can only let our registered students see it!