How can constructivist principles be applied to mass Higher Education? Can computer-based media enable it? Constructivism implies an environment with learner control, authentic learning contexts and diverse social interactions. The constraints of mass higher education make this impossible by traditional methods but computer-based media may enable it. A new semester course about the Internet for 300 students provided an opportunity to attempt such a course design, using the World Wide Web, email, video, and other media. The case study describes the course design based on constructivist principles, its implementation and evaluation.

‘Within the last few years, a distinctly different body of theory has begun to attract the attention of the field of instructional design and development -- constructivism.’ (Knuth and Cunningham 1993). This is not a novel theory springing complete onto the educational scene, but a developing body of related concepts and values with long historical roots (Entwistle et al. 1993). Constructivism is based on the view that knowledge must be constructed within the cognitive structure of every individual, so that it is fundamentally personal, while being dependent on experiences in the learning environment and on social interactions. It therefore stresses active learning processes, authentic learning contexts and assessment, and collaborative learning (Grabinger and Dunlap 1995, Lebow 1993). It can be contrasted with objectivism, the traditional view that knowledge is an external entity with an absolute value which can therefore be transferred from teacher to learner (CTGV 1993, Duffy and Jonassen 1992, Clayden et al. 1994). Both behaviourism and some cognitive psychology is based on the objectivist view that knowledge (meaning, understanding) has absolute, external values so that learning during instruction involves its transfer from experts to novices.

It has been argued that constructivism is particularly relevant to higher education (Entwistle 1993, Jonassen et al. 1993) but the conditions there do not initially seem favourable for its implementation. The reality of mass higher education is often limited resources, increasing student/staff ratios, increasing diversity of student types and ‘legacy systems’ of time-tabling and assessment. How can constructivist principles be applied to mass Higher Education? Can computer-based media enable it? (Duffy and Jonassen 1992, Jonassen, Mayes and McAleese 1993)

The opportunity to address these questions arose when the author designed and taught a new, one-semester course about the Internet and cyberspace for nearly 300 undergraduates doing Arts and Social Science degrees. As part of Keele University’s multi-disciplinary philosophy these students also take a two semester ‘Subsidiary’ course in a Natural Science. Many of them choose ‘Computer Science’, which for most involves a first semester course giving an introduction to IT and then one of six options in the second semester, of which this new course was much the most popular.

The experience and interests of this audience suggested that the course aims should combine some essential technical skills, their application to other subjects and wider social effects. The course learning objectives were that students would to be able

The pre-requirements were a working knowledge of email and Windows. All students have computer registration, e-mail and Internet access.

This case study describes the design of what was intended to be a rich environment for constructivist learning to achieve the learning objectives despite the constraints, largely by the use of computer media. Can computer based learning bridge the gap between the need for constructive, collaborative learning and the constraints imposed by large student numbers, limited resources, time-tabling and assessment requirements? The course used a mix of the World Wide Web, local Computer Assisted Learning (CAL), video, email and assessment methods. The constraints, the design, implementation and evaluation are described below.

An initial questionnaire asked about background and preferences for learning activities. Of 297 students registered, 255 were active of whom 229 completed an initial questionnaire, after an initial lecture. Their background is described here; answers to questions repeated in a final questionnaire are described later.

Of the respondents, 56% were female. The median age was 19 and 63% were either 18 or 19 years; 17% were over 21 years and the oldest was 59. 86% had British Nationality and of the rest Germans and Greeks were the commonest.

The students’ main subjects were a wide range of humanities and social sciences: management, English, modern languages, international relations, history, law and criminology were common. Only 18 studied a natural science.

On their computing and educational experience, most had completed the introductory IT module in the previous semester; all should have used email. 46% had used a Web browser before and 44% said they had used some Internet searching tool, but few had used other tools like ftp (6%) or Usenet (9%).

On their experience of learning methods, 40% had done assessed work in pairs or groups before but only 13% had kept a diary or journal, which may explain its poor completion. 48% had previously been asked to comment on their own learning processes. To an open question for comments, 17% included expressions of anxiety and hope, preferences for the use of paper and personal contact, and comments about the first lecture.

Grabinger and Dunlap (1995) have summarised the literature on constructivist learning theory and the five instructional design it implies, which they term a ‘rich environment for active learning’ (REAL). Constructivism describes learning as active knowledge construction by learners, knowledge acquisition in realistic contexts and the social negotiation of learning. These imply a need for learning environments characterised by five themes:

Lebow (1993), Jonassen, Mayes and McAleese (1993) and Simons (1993) have similar lists. Each theme will be described with its implications for the course design. Because the themes overlap and reinforce each other, one course element can be related to more than one theme, and they are listed under the most appropriate theme.

Learning should be student-centred in the sense that students take initiative and responsibility for their own learning. Learning should be purposeful, which implies that its aims are under student control. It should be self-regulated which involves self-questioning, reflection on learning processes and ‘metacognition’ - learning to learn. These features resemble those described for adult learners (Ference and Vockell 1994).

There are several implications for course design. Firstly, students cannot feel ownership if the learning goals and methods are strictly defined by the tutor. There must be room for negotiation with the tutor on the content and methods of learning. Assessment was through submission of seven practical tasks, a course diary and a final course report. Practical instructions concentrated on the tasks to be accomplished rather than the keys to be pressed. While the topics of the practical sessions were fixed, some practical tasks allowed individual choice: for example, the practical task on Internet searching allowed students to chose their own research subject; they had a free hand in creating their own home Web page. The course therefore allowed two alternative reports structures: a research report on Internet resources found for a subject, or an essay. Both the subject for research and the title of the essay were negotiated with the tutor (by email).

Secondly, students should have the opportunity to manage their own time. This requires the course to be resource-based, but a schedule was provided as a framework to encourage steady progress. All resources were available outside scheduled sessions, mostly through the Web. Tasks were required to be submitted within a week of the schedule although extensions were granted.

Thirdly, to make metacognition explicit, students should assess their own learning. Self assessment could take place when completing questionnaires during the course - initial and final ones and an intermediate ones on essential skills and progress in the course. Students were required for assessment to keep a diary of learning activities week by week, and asked to comment on their learning.

Active construction of knowledge requires that learning involves the generation of something concrete by the use of knowledge and the practice of skills. Typically this involves problem-solving, investigating and researching, to create solutions to authentic problems which the learners own. It was made clear in the introductory lecture (and a Web page) that the course had a hierarchy of learning objectives: knowing facts (verbal information, terms), using facts in skills with software tools, using skills to achieve (assessed) goals, reflecting on issues created by these capabilities and reflecting on their learning about facts, skills and issues to improve their own learning methods.

The role of the tutor in such learning is not the ‘the sage on the stage’ but a ‘the guide on the side’ (Ackermann 1996). One description of such teaching and learning is cognitive apprenticeship (Collins, Brown and Holum 1991) which tries to make visible the thinking processes involved in using cognitive skills. Techniques involved include the tutor thinking aloud as they perform a task and then supporting the learner as they do it, gradually removing the support.

Amongst the software tools needed to support a constructivist learning environment, hypermedia is often pre-eminent (e.g. Jonassen, Mayes and McAleese 1993, Knuth and Cunningham 1993, Spiro and Jeng 1990, Brockenbrough and Hoffman 1993). Hypertext can be an key tool for providing information through task-driven searching and the World Wide Web has many well known advantages as a hypermedium. In this course, resource pages provided links to resources of many types both local and global. Two other tools allowing students to represent their knowledge explicitly were word processors and a semantic net editor.

Generative activities included the practical tasks, which generated specific results, for example, a file retrieved or a confirmation of an email list joined. These were submitted as evidence of a skill being used. Other tasks involved producing learning materials: for example, a summary of the contents of a CAL package, and a Usenet contribution recommending a favourite Web site. Searching and evaluating Internet resources on a subject of individual choice were amongst the practical tasks and were also an option for the course report.

Another explicit encouragement for metacognitive activities was the task of drawing an individual concept map of cyberspace. Extensive, paper concept maps of the domain were available in the lab. The Courseworks concept map editor (Miller 1995) was available to create an individual map and a first stage document of some unlinked concepts was provided. Paper maps were allowed if the use of Courseworks proved too difficult in the time allowed.

An ideal learning environment would be rich in personal interaction with tutors and other learners. Interaction with a tutor is necessary for the first two themes but face-to-face tutorials or individual negotiation with each of 300 individuals were impossible. Three features of the course addressed this major problem.

• The tutor was available before and after the weekly lecture/video and informally in practical sessions when time allowed.
• Secondly, regular contact with the tutor was maintained through email and Web forms. The tutor regularly emailed all students using a course alias and reading email was the first student activity in practical sessions. In mid-course students were invited to submit outlines of course reports to the tutor by email for formative evaluation, returned by email. Individual students with particular problems emailed the tutor, using a special address.
• More routine feedback from students was provided by 5 questionnaires completed as ‘Web forms’ (documents with buttons for choices and text boxes for written text) presented as part of the practical work. Their contents were returned automatically to the tutor as e-mail. An initial and final questionnaire are considered in detail below. One mid-course questionnaire inquired about skills and progress to date. Another asked for choices on options and subjects for the course report; by default these were acceptable unless an email response suggested an improvement. Another questionnaire asked for an evaluation of the CAL package used at the start of the course.
• Personal contact depended largely on the Demonstrators provided in practical classes (at 15/1 or better) and in open access ‘surgeries’ every lunch time. Demonstrators were briefed before the course and an email alias was used to keep them informed of problems and changes during the course.

Learning experiences should be as realistic as possible, faithful to the original phenomenon, in contrast to abstract descriptions or ‘inert knowledge’. This reinforces elements of some other themes. Realistic problems allow students to address them realistically, taking ownership of their solutions (theme 1), developing deeper, richer knowledge structures. Such real, complex problems require more systematic problem solving methods (2) and are also more likely to benefit from collaborative efforts (5). This includes models of learning such as situated cognition (Entwistle et al. 1993) and cognitive apprenticeship (Honebein et al. 1993, Collins et al. 1991).

One way of providing such instruction is to anchor it in a real-world problem, event or issue which may be appealing and meaningful to students. The course addressed this in two ways. The practical tasks allowed students full access to the Internet (as opposed to a course Intranet designed for the purpose). This caused some additional problems such as slow response times, missing Web pages and so on, characteristic of Internet. Some important documents were made available as local copies, with the authors’ permissions, to ensure availability. Full Internet access meant that students had to behave responsibly in Usenet groups, e-mail lists and so on, and this responsibility was made clear to them in the first lecture.

The other way of providing realism was to use video clips extensively in the lecture periods. Video combines dramatic power in presenting real people, events, organisations and case studies, with a depth of detail which does some justice to the complexity of real-world issues. Bransford et al. (1988, 1990) proposed the use of video to mediate between the real world and the classroom, and to provide a case-based approach to teaching competencies. After the first lecture introducing the course philosophy and administration, the lecture periods were used to show video clips compiled by the tutor around issues relating the Internet to education, business organisation, democracy, freedom of speech and censorship, employment and gender. The videos were also available for viewing in the Library and text summaries were available as Web pages. The video material was particularly relevant for the assessed essays.

On award-bearing courses, the major motivation for students is assessment for certification. Therefore, inappropriate assessment will undermine an otherwise well designed course. The first, obvious, point is that assessment must test the learning objectives. In particular, the assessment of skills must involve using the skills not describing them verbally (Gagne 1985). Secondly, assessment must be authentic (Wiggins 1989): realistic in complexity, requiring students to contextualise their knowledge, requiring knowledge in depth rather than breadth, and diverse in form to allow for students’ differing intelligences and strengths. Thirdly, students must be told of criteria for the assessment of good performance in advance of assessment, and ideally at the start of a course. Assessment criteria thus not only reveal achievement to the examiner but also reveal the standards of the domain to the student.

The assessment of the module which this course replaced had been through practical exercises which were graded and a two hour examination by multiple choice questions. Multiple choice encourages shallow, broad learning at odds with constructivist themes. The examination was replaced by a final course report of 1500 words minimum marked by the tutor. Although creating a higher workload, this was necessary if the assessment was not to subvert the rest of the course design. It was required on paper to simplify marking and moderation in a short time. The assessment philosophy, criteria and administration were described in the first lecture and a Web page. Marks for the report were moderated by a colleague.

A personal diary of activities dyring the course was required, of a minimum 500 words. This was an opportunity to record and reflect on their learning.

The coursework assessment was retained but modified. During the 11 practical sessions, seven varied tasks were performed for assessment. All generated results, usually text, which were submitted to a special email address. The tasks were not intended to be demanding for the majority of students; they demonstrated basic skills. They were marked as excellent, pass or fail (if the task had not been performed). Although not all tasks allowed creativity, deep understanding of each topic or a demonstration of individual strengths, as Wiggins (1989) recommends, at least they were public, not unduly time-limited in execution, enabling of new skills, progressive in complexity and 'criterion-referenced' in measuring objectively the achievement of learning objectives.

Collaboration with fellow students can have several benefits to learning (Slavin 1991). Students can encounter different points of view which may clarify misconceptions and ineffective solutions to problems and give rise to synergistic insights. Group members must understand their different roles and learn to accommodate conflicting ideas. This reinforces individual accountability and responsibility (theme 1) and has been shown to give positive benefits to learning the subject matter.

The difficulties for group learning were these:

Within these constraints the following steps were taken to encourage collaboration.

This design for a REAL, about a complex domain yet within the constraints, can be summarised as authentic assessment of learning objectives, a resource base and media for personal interactions.

The course assessment, following from the learning objectives, replaced a two hour multiple choice examination with a final report of a choice of types and on a negotiated topic. This imposed a considerable marking load on the tutor, but one judged as being necessary. In contrast, the assessment of practical tasks were simplified to the level of a demonstration of basic skills. A personal diary of their course activities was required, providing an opportunity to record and reflect on them. Practical tasks generated a result which demonstrated a basic skills, and some produced knowledge products (summaries, concept maps).

The resource base was provided through the Web and videotape. The advantages of videotape were described above. The Web was chosen as the medium for course resources for the following reasons.

• Web forms allow feedback from students through questionnaires in a simple way which became routine.

Disadvantages of the Web compared to paper were

• it can't be annotated by students
• it is read from a computer screen, with lower contrast, unless printed
• access through the network can be slow.

The Web pages provided were

A variety of personal interactions are required for most constructivist themes. The importance of such interactions in education has been described in many ways: as feedback (Opheikens 1992), dialogue (King and Honeybone 1996), learning conversations (Harri-Augstein and Thomas 1991) and discussion (Paterson 1970). Face-to-face contacts were limited by timetable and student numbers. The solution used was to substitute some interaction with CAL, and use Web forms, email and the Usenet conference for interactions with the tutor and Demonstrators.

There was regular use of e-mail. An alias was used to send administrative messages and recent developments to all students, and another for Demonstrators allowed a private discussion of these in parallel. Students sent e-mail to the tutor and to Demonstrators concerning assessed work. They submitted practical tasks to special e-mail addresses, which returned a receipt for their work. Email was used as a backup if the Keele Web server failed (as it sometimes did on Monday mornings).

Prior to Web-based CAL being used, in the first two weeks a local CAL application provided an initial overview of the course while students were familiarising themselves with course administration, software and off-campus resources (and re-familiarising themselves with Windows and email). It was developed using the Toolbook template from the CTICM (Price 1995). Required minimal skills to use, it had 32 screens of text, graphics, and quizzes. As a first task, students made a summary of its contents in the Notepad as they used it.

One complaint (at a mid-course student liaison meeting) was reading the practical instructions on the screen while performing other tasks. (It was possible to print Web pages a day in advance but few students did so.) So the current week’s instructions were projected on a screen in the computer laboratory.

Completed as a Web form, a final questionnaire repeated much of the initial questionnaire to detect changes during the course. Web forms required students to include their username to allow cross-tabulation between questionnaires. Forms requested a password (distributed by email) to detect spurious submissions from outside the course (there were none). To get further information some students were interviewed in the last two weeks of the course. The mid-course questionnaires and student assessments are also briefly described below.

The 159 responses to the final questionnaire were made in the last week of the course. Of these only 133 completed both initial and final questionnaires and could therefore be used in the comparisons of the same students below. The results for these 133 replies were similar to those of the total initial and final questionnaire responses.

In the first week 42% expected the course to be difficult, only 3% to be easy. By the end some 10% thought the course easier than expected. Half those who thought it was going to be difficult later thought it was only average.

The significance of differences between the initial and final questionnaire results were tested by a non-parametric, two-tailed Sign Test using the three answer categories as ordinal data. A shift towards thinking the course easy or average was significant (p<0.001).

At start of the course 63% expected to enjoy the course. These expectations were somewhat disappointed; at the end 27% had enjoyed the course, 17% had not, while a majority enjoyed parts. There was a significant change (p<0.001). Of the 45 students who initially did not know, 28 enjoyed it somewhat and 11 did not.

It is puzzling why significantly fewer students interviewed said they enjoyed the course than those that thought they would. This may be due to unrealistic expectations of ‘surfing the net’ generated by recent media coverage. The course showed the Internet as complex and many-sided so that the reality disappointed.

At the start, of those completing both questionnaires, 22% preferred only lectures, 23% only CAL. Though 68 students changed their response, there was no overall change in attitudes (p=0.9).

Initially, 20% preferred lectures only and 26% videos. A small change in favour of videos was not significant (p=0.21) and the pattern in the full 159 completing the final questionnaire was the same.

Initially, 43% wanted assessed work done in groups, only 17% were against it, but this changed so that more wanted individual work (p<0.001). (Only 24 of 255 did reports in pairs!)

Initially, 77% preferred coursework only rather than an examination alone or in combination. At the end, having completed the coursework and much of the course report, 92% wanted coursework alone, a significant change (p<0.001). Students had examinations for two other subjects at the end of the semester and did not want another.

Initially there was a 53% preference for paper against 13% for online documents. There was a shift from paper to online documents (p=0.01).

Initially only 14% preferred submission of work by email with 56% against. After the course, fewer were undecided and 34% were in favour of email submission. The change is significant (p<0.008).

In week three a questionnaire evaluated the use of the introductory CAL package. Students made a summary (in a editor) while reading it. The median time for making notes was 4 hours (with a range from 1 to over 4.5 hours) which was longer than the 3 hours intended. Nonetheless, of the 189 responses, 69% preferred the CAL package to a lecture for making notes, although only 42% thought they learnt more by making notes than receiving printed copies (39% thought not).

In week six a questionnaire asked about initial skills and their current progress (appendix 3). Of 205 respondents, 28% said that they did not have enough skills in email at the start of the course, and 20% said the same of Windows. Asked if they had been anxious about the adequacy of their skills, 40% replied that they had been at the start and 23% that they still were.

Asked about printing Web pages (hard copy appeared in the Library building for collection), only 27% said they had printed a few pages. 59% said that they were up-to-date with practical work, and only 11% were one week behind or worse and this included several starting late.

In week seven a questionnaire asked for a choice of their course report, between an essay or research into Internet resources, and a suggested title. Submissions which were unsatisfactory received email replies. The option to work in a pair was selected by 12 pairs.

In the last two weeks of the course 56 students were chosen at random, by seat position in the computer laboratory, and interviewed by the author sitting next to them. A draft of questions was piloted and the final list of 40 questions was completed by the author during the interview. Most were multiple choice with three options.

Of the 56 students, 41 described the course as being of average difficulty, with a few thinking it easy or thinking it difficult, fewer than thought it difficult in the final questionnaire. Particular difficulties included inexperience with IT generally, with email, and with Windows operations, fear of the technology, difficulty understanding new terminology and understanding the practical instructions, having to read significant amounts in general and from the screen in particular, and general problems with English in non-native speakers. The factors making it easy were fewer: having had significant IT experience, the course being largely practical and having all the information available on the screen were mentioned.

Half of the interviewed students said they had enjoyed the course and only three said they had not. This is more positive than in the final questionnaire (p<0.001) : of the 35 who answered both 17 were more positive in the interview. The questionnaires are likely to be more honest.

The aspects of the course described as enjoyable were that it was largely practical, producing a Web page, searching and browsing the Web, using email, feeling confident about the Internet, the interesting subject, the videos, finding information relevant to other courses, that it was well structured, that help was available, concept mapping, that they had prior experience and they found it easy or fun. The parts of the course mentioned as un-enjoyable were the amount of reading from the screen, frustration, difficult content, fear of the technology, the introduction was too slow, the videos, the CAL, searching for information, concept mapping, and the Demonstrators. In short, most parts of the course were either liked or disliked by some students, except that complaints about the amount of reading from the screen were common, and many people enjoyed creating their own Web page. The topics mentioned as most useful were searching methods, finding information on other subjects, ‘navigating’ the Internet, creating a Web page, email lists, ftp, and telnet. Replies to other questions are described below.

These replies are consistent with head-counts in lectures. Lecture attendance declined exponentially from over 220 in the first week down to the 30’s. A significant minority used the library copies of videos.

Did you find the practicals enjoyable?

Have you missed any of them?

Have the practicals given realistic problems to solve?

Practicals were as enjoyable as the course overall, attendance was good and the tasks were thought realistic.

Have the demonstrators been supportive of learning?

Have you come to lunch time surgeries?

Have you emailed the demonstrators or tutor for help?

If so, did it work?

The personal interactions provided were appreciated. An open ended question asked how the demonstrators might be more useful. 36 students could make no suggestions for improvement. Others commented that more were needed, that they could be more knowledgeable and that they could explain more clearly. Surgeries and email were used for help by a minority, to good effect.

Have you printed many Web pages

Would you print more if it were easier than collecting copy at the Library?

Have the projected instructions in the laboratory been useful?

Have you used the WWW outside the practical classes?

Have you had more freedom over the pace and method of work in this course than other Keele courses?

Did you enjoy that amount of freedom and responsibility?

Most students liked the practical instructions being displayed in the computer laboratory rather than reading them only from the screen. Despite the preference for paper materials, only a minority printed many Web pages even if it were made easier to print them on the spot. The use of the Web outside practicals emphasised the value of accessibility through the Web. Most students said they had more freedom in their learning than in other courses, and they liked that.

Have you talked about the course with fellow students?

Have you used the course Usenet group (apart from the required task)

Have you done your report in a pair?

Did the first lecture encourage you to discuss the course with other students?

Most students talked about the course with others, not necessarily on the course, but few tried the Usenet group. In fact there were 290 student messages in the Usenet group. With students being required to add one message as a practical task, only about 35 messages were added proactively, many from a small number of individuals. About a third of students thought that the introductory lecture had encouraged them to discuss the course in practicals, Usenet or by email.

Out of 255 reports submitted, only 24 worked in pairs. Reasons given for not working in a pair were that they preferred working alone, it was easier or more convenient or faster to work alone, they could choose their own topic, that they did not know others to work with, that the partner they were working with left the course. Reasons for working in a pair were that discussion was good, that they could do a more thorough job, that it was more interesting, that it reduced the workload, and that they worked well together.

Assessment.

Assessment methods

The assessment criteria were

Assessment tasks test useful skills

Do you like choices for the report

Clear about writing course report

Have you kept a diary

Did you find the practical tasks

Did you get the tasks done on time

Are you going to write about your own learning processes in the report?

Are you happy to write about that?

For most students the assessments tested useful skills, were varied enough and the criteria were clear. All liked the choices for the course report. One area of doubt was the diary, where only two thirds said they had kept one fully. In the submitted reports the absence of a diary was the commonest problem, with 14 students not submitting one. A quarter of the students did not intend to write in their reports about their own learning processes, and did not wish to. The students were equally divided on the difficulty of the practical tasks but most had submitted them on time, as confirmed by the questionnaire in week 6.

A variety of answers were given to an open question for comments to improve the course. Leaving aside administrative points and general complaints of difficulty, some students wanted: less reading especially from the screen, paper notes or easier printing, more group work, more use of the Web. clearer instructions, help with the diary, want clearer instructions for assessment, more demonstrators, and more freedom in learning methods. Others said that it had been an interesting, relaxed or great course. Comments in the diaries were varied, from the laudatory (‘I really enjoyed this course, it gave me new skills for the rest of my life’) to a variety of specific dislikes and criticisms.

To pass the course students were required to complete 8 practical tasks and submit a satisfactory course report. Of the 255 reports, only 31 described research into Internet resources, in a wide range of topics. The rest were essays mostly on issues described in the videos.

Of 296 usernames appearing on the course, 41 submitted no report and attended few or no practical sessions and either formerly or effectively withdrew from the course (and possibly the University). It is unlikely that the withdrawals were related to the course design. Of the 255 active students, 208 passed (82%); 18 of these were given A grades and 23 were marginal passes. The pass rate is low compared to other options, but the course was much the most popular option and it replaced a course which in previous years had been regarded as easy. For many students it was the default option, and their motivation and application may have been lower than average.

In what ways was the course design successful in creating a rich environment for constructivist? For each theme we consider the outcome and the students attitudes, and some possible improvements.

Did students like the opportunities for responsibility over content and methods? 64% of interviewees said they had had more freedom in this course than in others, and almost all liked that. They all liked the variety of assessment methods although 14% found the assessment instructions unclear and 20% were unclear about writing their report. While the first lecture, a Web page and a reminder email spelt out the choices for the assessment, this added a complexity with which some students had not come to terms.

Did they manage their time? One of the mid-course questionnaires asked whether students were up to date with the practical work and 90% were. A similar proportion of interviewees had submitted some or all tasks on time. 10% were late in submitting final reports. So the majority of students did manage their time to the deadlines.

Did they comment on their own learning processes? Of the 56 interviewed, 32% did not confirm they had kept a diary and 5% did not submit one. Only 35% of diaries included comments on their learning or the course (the others just presented a record of activity). 60% of interviewees said they were happy to write about their own learning but half of them did not do. This indicates that more guidance is needed on how to write the diary.

The practical sessions were well attended and the tasks were well completed. About 90% of interviewees thought that some or all tasks were realistic problems, and nearly all enjoyed at least some of them.

The types of software supporting generative activities (apart from text editors and word processors used for all written work) were the Web as a hypertext resource, wordprocessors, the CAL package and the concept mapping package Courseworks. About half had used the Web prior to the course and 71% of interviewees were using the Web outside the practical sessions. Comments about using the Web were generally positive, especially creating a home page. The common complaint was having too much reading from the screen. Nonetheless, there was a shift in attitude of about 10% towards online documents rather than paper, and a bigger shift towards email rather than paper submission of coursework. There was surprisingly little printing of Web pages: 70% of interviewees had printed no course Web pages, and only 9% would definitely print more if a printer was more accessible. Projecting the relevant Web instructions on a screen in the PC lab was widely appreciated. In future the option of printing pages in advance of practical sessions should be stressed and the course pages will be restructured to make this easier.

The questionnaire evaluating the introductory CAL package showed that 69% preferred the CAL package to a lecture, although many would have preferred to receive the notes than to make them themselves, but this would hardly be a generative learning activity! There was no overall change in attitudes to CAL.

As the Courseworks concept mapper was used in the last practical session, its use was not systematically evaluated. Unprompted comments were predominantly positive. In future it will be introduced earlier in the course so that maps can be built over a period.

Another generative activity which was encouraged but not required was the email submission of an outline of the report for formative evaluation but only 3 were submitted, although there were a score of other enquiries about reports.

This aspect of the course was successful within the technical and administrative constraints. Using the full Internet, rather than only an Intranet, produced some technical problems. Transmission times from the US in the afternoons were slow. The TONIC web-based CAL course had a few glitches which could only be reported and worked round. No incidents of abuse of Internet access occurred.

Attendance at the videos declined exponentially to about 10%. The scheduling was regretted by all and about 20% intended to use the Library copies. Improvement in this area depends on the timetable. However, more integration of the video content could improve attendance, for example asking for questions relating to the issues in other tasks.

Student views on the course assessment were generally positive. Preference for coursework over examination was strong. The choices of report subjects was also appreciated. Nearly all interviewees thought the assessment methods appropriately varied.

The pass rate of 82% was disappointing, despite a significant shift towards thinking the course easier than expected. Producing work for this Subsidiary course took second place to their main subjects. A third of the failures were due to not submitting a diary (only 13% had used this assessment method before) and this can be improved by more prominent guidance.

At the start of the course, the prospects for group work looked fair; 40% had previously done assessed work in groups and 43% definitely wanted it. But the timetable and physical layout of the laboratory was unhelpful and there was a significant drop in preference for groupwork in favour of individual assessed work. Only 9% wrote their reports in pairs.

At an informal, non-assessed level, three-quarters of interviewees had talked about the course with others but use of the Usenet course conference was low. Only 28% of interviewees thought that the encouragement to collaboration had affected their co-operative behaviour.

Co-operation was not specifically rewarded by the assessment methods. The constructivist arguments for co-operation are not driven by reward structures, but by the value of dialogue in knowledge construction through critical questioning of their own and others’ views ( King and Honeybone 1996). In his review of co-operative learning in education, Slavin (1991) describes the commonest designs (student team learning) which are driven by the reward structure: ‘It is not enough simply to students to work together. They must have a reason to take one anther’s achievement seriously.’ To this end, the commonest reported groupwork studies use rewards for teams which require all its members to learn, if necessary by teaching each other. Kohn (1991) criticises this emphasis on reward systems to persuade students to collaborate as ultimately counter-productive. Instead he proposes ‘a carefully structured co-operative environment that offers challenging tasks, that allows students to make key decisions about how to perform those tasks, and that emphasises the value of helping each other learn’ as an alternative to extrinsic motivators which is ‘more consistent with the ideals of educators’. While compulsory groupwork supported by a reward system could have been used (and was used in the previous course), enforcing collaboration in this easy way is inconsistent with constructivist philosophy. Methods of encouraging co-operation without coercion are still being considered.

Although the ‘median type’ of student was a 19 year female Humanities student with some IT experience, the most notable feature of the student body, in the statistics and the flesh, was their diversity. Few answers were unanimous and they were often evenly divided. This diversity is a challenge for mass Higher Education in general but constructivist designs are helpful: encouraging student responsibility and initiative accommodates the variety of experience and attitudes.

Entwistle et al. (1993) concluded from empirical data that ‘the assessment procedures have a profound effect on the way in which students learn. Providing a constructivist teaching environment will have little effect on the quality of learning while conventional assessment procedures remain in place. The principle of constructivism will have to be applied to the design of new forms of assessment, perhaps relying much more on self- and peer-assessment than in the past.’ In this course a multiple choice examination could not fairly test the learning objectives, and the assessment methods used went some way towards a fair test. More and better self assessment in the diary is a further step.

Entwistle et al. (1993) also provided a conceptual model of the factors affecting teaching and learning in higher education. Three major factors were instructional methods (lectures, CAL, resource-based learning), assessment procedures and support from tutors, seminars and peer discussion. These were the three areas of course design here: the resource base, assessment, and personal interactions. In this respect at least, the design features are adequately addressing the needs for constructivist learning.

A final point is the interpretation of ‘constructivism’ and therefore how it should be implemented. Constructivism has its critics (Gruender 1996, Olssen 1996) but even amongst its supporters there is a range from mild to strong constructivism (Lowyck and Elen 1993, Merril 1991). Entwistle at al. (1993, p. 354) claim that there are good reasons from empirical evidence that learning environements should not be entirely (strongly) constructivist. In this course, as in their research, different students reacted quite differrently to the same features of the learning environment. There are different styles of individual learning varying along continua from holistic to serialistic, for example. In the same way some students will enjoy the challenge of constructivist learning while others will sometimes find them uncomfortable and need more objectivist instruction. A radically constructivist course would be more difficult to implement within the constraints of resources and institutional culture, so it is cheering to think that a partial implementation of constructivist principles may actually be optimal for the majority of students.

(tables tidied up May 2005)

Ackermann, E. 1996. Interactive tools: zappers’ paradise, creators’ nightmare. Keynote paper at Mediactive conference, Liverpool.

King, A and Honeybone A. 1996 Needs before means: the dialectics of learning and technology. Alt-J (Association for learning technology journal) 4 (2), 4-16

Miller, P. 1995 Courseworks: an extended concept mapping tool. Active Learning, 3, December, 50-53.

Opheikens, B. 1992. Telecommunications in distance education: user demands and technological possibilities. In Learning Technology in the European Communities. Proceedings of the Delta conference, The Hague, October 1990. Kluwer Academic Publishers, Dordrecht.

Paterson, R.W.K. 1970 The concept of discussion: a philosophical approach. Studies in Adult Education 2 (1), 28-50.