In a series of four studies, student postings on newsgroups created for their courses at Carleton University were monitored, and opinions were gathered from samples of students and instructors regarding their newsgroup activities. Results show that an overwhelming majority of students never posted messages on newsgroups, nor did their instructors. In addition, a large majority of students rarely read what others had posted. Students were more likely to post messages when graded for doing so and when encouraged by the active involvement of their instructors, though the quality of the resulting discussions remained questionable. The few instructors who were enthusiastic about the educational potential of course newsgroups worried about the large amounts of time spent on reading, answering, and grading students' postings and about the lack of career rewards for these activities. The results clearly indicate that course newsgroups have limited educational value until students and instructors find more time and motivation to make more and better use of the medium.

Contents

Introduction
Study 1: General Observations of CHAT Course Newsgroup Use
Study 2: Content Analyses of Active Course Newsgroups
Study 3: Student Survey
Study 4: Interviews with Students and Professors
Discussion

Introduction

It is now fashionable for university administrators to embrace computer-mediated communication as part of their strategic plans, in large part because the medium seems to hold great potential for increasing enrollments and decreasing costs. It also may hold some potential for education. Yet little is known about how much of the educational potential of computer-mediated communication can be realized. Optimists can envision thousands of happy students, credit card in hand, sitting at home in front of their Internet links to course Web sites and newsgroups, sopping up knowledge as fast as it comes online (e.g., Brown, 2000). Skeptics can envision an equal number of angry students working off credit card debts, seeing no educational or competitive advantage gained from an online McDegree in a postgraduate world (e.g., Jaffe, 1998; Noble, 1997). Between these extremes are several contingent possibilities that different features of the Internet might be pedagogically useful for different combinations of students, course topics, and learning objectives. One such Internet feature is the newsgroup, an organized series of messages posted to a public site that allow several individuals, such as students in the same course, to engage in an asynchronous discussion. The primary purposes of the present research were to determine (1) how course newsgroups are used; and, (2) under what conditions they might be useful, by examining course newsgroup use in one of Canada's most ambitious efforts to supplement traditional university teaching with newsgroup discussions.

The Carleton Hotline for Administration and Teaching, or CHAT at http://www.chat.carleton.ca, was established in 1993 as a pilot project to allow discussions among students at Carleton University (undergraduate enrolment about 15,000) in Ottawa, Canada (metropolitan population about one million) (see Mallett, Miller & Stewart, 1993). In its first few years of operation, CHAT administrators established a course newsgroup only at the request of a professor; by 1995 about 400 courses had one. An increasing desire of senior administrators to recast the university's image as "Canada's High Tech U", CHAT's proclaimed "user friendliness", a 1996 award of excellence for CHAT from the Canadian Association of University Business Officers, and funding from ScotiaBank, Canadian Pacific, and Bell Canada, promoted rapid expansion. Beginning in 1997, CHAT course newsgroups were automatically established for all 2,500+ courses and sections offered on campus. Articles about CHAT, including testimonials and advertisements for workshops on how to use it in the classroom, appeared almost weekly in the campus staff newspaper. Every student received a CHAT account for e-mail, Web browsing, and newsgroup activities as part of their tuition fees. Help for students wishing to use CHAT was well-coordinated and abundant. Such activities gave consistent indication that the administration had high hopes for promoting CHAT as an important educational technology.

There were some compelling rationalizations for these hopes. Recently resurrected ideas about dialogical (Socratic) and group learning (see Bruner, 1996; Vygodsky, 1986) remind us that students can learn by talking to each other, as well as by listening to their professors. However, such group work normally requires students to meet at the same time and place, an increasingly difficult requirement in a busy world. Computer-mediated communication eliminates the necessity for students to be at the same place at the same time, thus reducing two main barriers of group communication (Brent, 1999; Harasim, 1990; Kiesler, 1997; Schofield, 1994). However, the medium creates new requirements. For example, students must find a computer to participate in a newsgroup, which might be difficult if school computers are heavily used by other students, or if home computers are heavily used by other family members. Students must also find time and motivation to read and write messages in newsgroup discussions, which might be difficult if written messages are boring, if writing skills are lacking, or if priority is given to other time demands. It is questionable whether a technology that removes constraints of meeting coordination will itself lead to more discussion among students when the same technology generates new requirements of access, motivation, and time. Fortunately, the question can be addressed, perhaps answered, by empirical research.

Research investigating the educational effects of computer communication in primary and intermediate schools shows mixed results, some indicating modest improvements in motivation and learning, others indicating none (e.g., see Boettcher, 1994; Clarke, 1985; Hiltz, 1990; Kulik, Kulik, & Cohen, 1980; McQuillan, 1994; Nakajima, 1994; Perkins, Schwartz, West, & Wiske, 1995). Curiously, there is almost no research investigating the effects of computer communication in higher education. In one of the few reported studies, Althaus (1997) compared the effect of computer discussions with face-to-face discussions on students' academic performance in an undergraduate sociology course. One-hundred thirty-four students in face-to-face discussion groups were invited to participate in computer-mediated discussion groups for 5% of their course mark; about 60% volunteered and 31% submitted at least one message per week. Those who participated regularly in computer discussion groups liked the experience, attended class more regularly and received somewhat higher marks than those who did not participate. Yet, in related studies, Hiltz (1990) did not find an increase in exam performance, and Althaus acknowledges that the better marks and attendance in his study might be due to self-selection of students for online discussions rather than pedagogical benefits of the medium.

Despite the potential confounds of Althaus' study, two of his results seem especially interesting to us. First, 69% of Althaus' students did not regularly participate in online discussions despite the offer of a 5% bonus for their course mark (Althaus does not report how the grade for participation was assessed, but hints that it was based on quantity not quality). Second, though 94% of all students believed that computer-mediated discussions should be offered to all students, only 4% believed it should be required of all students and 90% believed it should be an optional exercise. The results suggest that, while it is difficult to condemn computer-mediated course discussions in principle, students in fact have other things to do and would not appreciate being coerced by grades to change their ways.

The suggestion is congruent with major principles of the economics of attention, first proposed several years ago (Thorngate, 1988, 1990a, 1990b, 1991, 1997; Thorngate & Plouffe, 1987). In brief, though information continues to grow, attention remains a fixed resource, simply because to "pay" attention is to "spend" time and nature does not expand lifetime to fill the information available. The result is an attentional economy in which attention is exchanged for the production and consumption of information and in which an abundance of information creates a buyer's market. A typical North American student is supposed to spend about 1,600 hours in class, and encouraged to spend perhaps three hours of study for each class hour, to receive the information required for a bachelor's degree. This is about 1% of an 80-year lifetime, averaging about 15 minutes a day or about the same amount a student will spend in a lifetime of showers. But attention spent on undergraduate university education is not distributed across a lifetime; it is instead normally concentrated in four 8-9 month periods. A typical 7x24 = 168-hour week of university might require 15 hours of lecture and 45 hours of study. Normal sleep might require 7x8 = 56 hours more. Many of the 52 remaining hours in a week must be spent preparing and eating meals (estimate 10 hours), attending to personal hygiene (4 hours), doing household and shopping chores (6 hours) and in transit (7 hours), leaving perhaps 25 hours in a week for other activities. About 60% of undergraduate students at our university have a part-time job, in part to pay for their tuition fees. About 10% of students have children. The remaining 25 hours are quickly consumed by such obligations, leaving very little time for other activities. Exams and paper deadlines are not distributed evenly across an academic term/semester, and when academic time demands are clumped, priorities are set or shifted, and many good intentions are ignored.

Computer-mediated discussion groups may be a terrific idea, but we suspect that for many students they represent yet another demand on time in short supply. It should thus not be surprising that a student would ask what benefits might accrue from participating in one, or in two, or in ten course newsgroups each academic year. At one extreme, a student might spend time checking for new entries in a newsgroup but discover than there are few, if any, posted and interesting. With no payoff for the time invested in checking, the time is likely to be diverted to more rewarding tasks. At the other extreme, the volume of newsgroup messages might be so high (consider a course with 500 students, each required to produce one message a day), that reading them all, much less responding to many, might take far more time than is available. Should a student spend 8 hours this week cramming for an exam counting 30% of the course grade, spend the same time working to pay the next tuition fee installment, or spend it reading and writing newsgroup entries for the intrinsic reward of peer-to-peer communication or to boost a minor class participation mark? And if a student misses three weeks of newsgroup activity, would it be possible for the student to ask for work time off or a three-week extension of a term paper deadline to catch up on the newsgroup postings? Considering such competing demands for a student's attention, we might expect that computer-mediated class discussions would often succumb to higher priorities, and that students would not wish to be forced by threat of a low grade to attend to the discussions when other time demands are pending.

Yet students might not be the only ones affected by the time demands of computer-mediated course discussions. Professors and teaching assistants who encourage computer-mediated communication are sure to be affected too. The most well-meaning instructor might not find time to organize or comment on all, or even many, student comments in course newsgroups, especially if the newsgroup generates considerable activity. Indeed, the result could be paradoxical: students might flock to courses newsgroups if their instructors promised personal comments on their postings which, in turn, would demand more attention of instructors, until the limits of instructors' time reduced their comment rate which, in turn might cause students to cease posting.

Our thinking about the time demands and constraints on students prompted us to develop four studies of CHAT use. The first study simply examined the distribution of messages posted in the 2,512 course newsgroups automatically begun by Carleton University computer staff for each course and section offered in the 1997/98 academic year. The second study examined a sample of 40 of these newsgroups, ranging from barely to very active, in order to determine frequency distributions of authors and content. The third study asked students in three courses about their use of CHAT, including their use and opinions about course newsgroups, in a standardized questionnaire. In the fourth study, 21 students and 20 professors, ranging from newsgroup-philes to -phobes, were interviewed in depth about their course newsgroup activities, experiences, and suggestions for improving the medium.

Study 1: General Observations of CHAT Course Newsgroup Use

The purpose of Study 1 was to provide a statistical profile of CHAT course newsgroup use and to determine the rate of course newsgroup activity in courses from different disciplines and years (1st = Introductory courses, 2nd, 3rd and 4th = Senior or Advanced, plus graduate courses). These statistical profiles were used to select 40 of the most active CHAT course newsgroups for more intensive observation in Study 2.

Methods

Sample

All 2,512 course newsgroups developed on CHAT for the Fall term and the full year courses of the 1997-98 academic year were observed. We followed the progress of all newsgroups during the 12-week Fall Term, September - December 1997. In addition to recording the course discipline and year, we also recorded whether a message was posted as (1) a topic or as (2) a reply to a topic following the standard dichotomous structure of newsgroup postings. The topic/reply distinction gave us an opportunity to calculate the number of replies per posting as a rough index of student involvement in dialogue (many replies per topic) versus monologue (all topics and no replies).

Data collection

After recording the catalogue number and title for each course, we thrice counted the accumulation of new topics and replies: first at the end of week 4, then at the end of week 8, and finally at the end of week 12. As a result, by the end of the Fall term, we had a large database of three topic counts and three reply counts for each course. From these data, accumulation rates of postings, topics, and replies could be calculated for each course and year from the beginning of the course to the end, and the resulting indicators of newsgroup activity could be related to each other.

Our data represented a snapshot of course newsgroup activity for Fall 1997. In order to determine if changes had occurred during the years since then, we obtained cumulative message counts for all courses (Fall, Winter, and full year) given in the entire 1999/2000 academic year, kindly provided by the Carleton University CHAT coordinator. These counts were organized by course calendar number and were collected automatically by the CHAT computer. They did not separate topics from replies, nor did they subdivide counts according to beginning, middle, or final month of each course. As a result, comparisons with the original data were limited. Yet they provided a rough check for increase or decrease in newsgroup participation rates.

Results and Discussion

Table 1 shows the percentage of newsgroups with 0, 1-9, 10-39 and 40+ postings. Though the percentage of courses with no postings shows a 12% drop between 1997 and 1999, it is still remarkable that the modal number of postings in newsgroups for both years is zero. About 8.6% of the 1997/98 newsgroups, and 10.6% of the 1999/2000 newsgroups accumulated only one posting; in almost all of these cases the posting was an advertisement for a used textbook, a course outline, a party or other social announcement. Even with a relaxed definition of an active newsgroup as one accumulating 10 or more postings in 12/24 weeks of half-year/full-year classes, Table 1 reveals that the vast majority of 1997/98 course newsgroups (93.6%) and 1999/2000 course newsgroups (88.4%) were not active.

Table 1: Percentage of Newsgroups Accumulating Various Numbers of Postings

Course newsgroups	Number of postings in newsgroups
	0	1-9	10-39	40+
Fall + Full-year courses, 1997/98 (n = 2,512)	78.4%	15.2%	4.4%	2.0%
Fall + Winter + Full-year courses, 1999/2000 (n = 4,185)	65.8%	22.0%	7.3%	4.9%

To be fair, the proportion of course newsgroups with 40+ postings more than doubled in the two years; indeed, the most active of the 2,512 newsgroups in 1997/98 accumulated 350 postings, most active two years later accumulated 1,209. Yet their quantity does not necessarily reflect their pedagogical quality. For example, 350 accumulated in a second year Computer Science course when students ran a newsgroup contest to see who would post the last posting. The 1,209 accumulated in a second year ESL course that required postings for writing practice.

Of the 547 newsgroups from the 1997/98 academic year with at least one posting, 40% were first year courses, 30% were second year, 18% third year, 8% fourth year, and 4% graduate courses. These proportions over-represent first and second year course offerings (with 12% and 13% of all courses offered) and under-represent the rest (22% 3rd year, 23% 4th year, and 30% graduate courses), and in part reflect Carleton University's emphasis on linking course newsgroups to the large introductory courses offered via instructional television. The 547 newsgroups came from all four faculties of Carleton University: 31% came from Arts, 42% from Social Sciences, 18% from Science, and 9% from Engineering. These proportions parallel those of all courses offered at Carleton in the four faculties: 32% from Arts, 35% from Social Sciences, 18% from Science, and 12% from Engineering.

During the 12 weeks of the Fall Term 1997 a total of 13,851 messages were posted in the 547 active newsgroups: 6,441 topics and 7,410 replies, or about 1.15 replies per topic. Of these, 4,374 messages (31.6% of the total) were posted in the first four weeks of the term: 2,404 topics and 1,970 replies (0.82 replies per topic). Another 4,362 (31.5%) were posted in during weeks 5-9: 1,948 topics and 2,414 replies (1.24 replies per topic). During weeks 9-12 the remaining 5,115 postings (36.9%) were added: 2,053 topics and 3,062 replies (1.49 replies per topic). There was, it seems, a slight increase in the rate of posting as the term progressed. In addition, the number of replies per topic increased slightly as the term progressed.

We are tempted to interpret the small increase in the average number of replies per posting as an indication that active course newsgroups were becoming more dialogical and less monological as the term progressed. Yet we must do so with caution. There were wide variations in the reply/topic ratio across courses. Our informal glance of 100 active course newsgroups revealed about 24% contained almost all topics and almost no replies. Yet further inspection indicated that about half of these showed reply-like content posted as topics. It seems that even though designers of newsgroup software make an important distinction between topics and replies, many students do not. When content is considered by the printed word rather than by topic/reply category, about half the active newsgroups show an average of at least one reply per topic. But the content of these postings are not necessarily related to the content of the course, as we report in Study 2.

Study 2: Content Analyses of Active Course Newsgroups

Study 2 was designed to examine in more detail the content and group dynamics of the 1997/98 course newsgroups. The primary purpose of this study was to search for reasons why some course newsgroups were more active than others. Included in the search were analyses of the relationship between student postings and class size, gender, year, and faculty as well as relationships between student postings, instructor postings, and course requirements. In addition, we examined the trajectory of course newsgroup discussions, noting, for example, the distribution of replies to each topic and how many questions were asked and answered. Finally, we examined the content of the discussions to estimate what proportion was related to the content of their courses.

Methods

Sample

Forty of the 1997/98 course newsgroups with at least minimal newsgroup activity were selected to be analysed. Twenty of the 40 newsgroups were selected at random from the 50 newsgroups with the highest number of postings (40 or more). The remaining 20 were selected at random from the 114 newsgroups with 10-39 postings. We obtained enrollment statistics for the 40 courses, then read the posted messages for their newsgroups. Each message was categorized to obtain the following:

1. sex of the author (male of female);
2. position of the author (student or instructor);
3. type of message (new topic or reply to existing topic)
• if a reply, the number of days between the reply posting and the posting of its parent topic
4. length of message in lines
5. content (questions, comments, course-related material, etc.)
6. style of writing (personal, factual, humorous, etc.)

We excluded from the sample all 383 course newsgroups with 1-9 postings simply because there was so little content in them to analyze.

Results and Discussion

Perhaps the most interesting finding of Study 2 is that only a small proportion of students posted messages in the course newsgroups. Of the total 4,256 students enrolled in the 40 courses, only 550 (13%) posted anything. On average, 53% of these 550 students were male and 47% were female, though males represented only 46% of the students in the 40 courses. Thus, males were slightly but reliably over-represented in the postings, t (35) = 2.48, p < .02. The small difference may reflect a residual sex bias of newsgroups as a "male" medium.

Surprisingly, there was no significant correlation between the course enrolment and the number of students posting in course newsgroups (r (38) = +0.005, p > .10), nor was there a significant correlation between course enrollment and the number of postings per student (r (38) = - .06, p > .10).

Something other than class size appeared to account for the variation in course newsgroup activity. One possibility is the instructors' activity. Perhaps instructors who demonstrated enthusiasm for their course newsgroups by posting many messages would lead by example and encourage students to contribute more. Perhaps not. The correlation between the number of instructor postings in the 40 courses and the percentage of students in the courses who posted at least once was insignificant, r (38) = +0.14, p > .10. So too were the number of instructor postings and the number of student postings across the 40 courses, r (38) = -0.10, p > .10, as well as the number of postings per active student, r (38) = -0.08, p > .10 respectively. In sum, there is no evidence from the results that instructors influence students by their posting activity.

At least two other alternatives remain: an instructor's verbal encouragement (e.g., "OK, students, log into your newsgroup to find some exciting discussions!"), and grades (e.g., "10% of your course mark will depend on the number and quality of your postings."). Following several phone calls, we managed to obtain information from 23 of the 40 instructors about their newsgroup grading policies and encouragement practices. Five of the 23 instructors graded students for their postings (5 - 25% of the course mark), eight provided verbal encouragement to students to post, and 10 provided neither. The effects on student participation are striking. In the five graded courses, an average of 81% of students posted at least one message. In the eight courses with verbal encouragement only, an average of 31% of the students posted at least one message. In the remaining 10 courses, an average of only 11% of the students posted. A one-way analysis of variance showed these three averages to be significantly different, F (2, 20) = 14.4, p < .001. It thus appears that newsgroups are rarely sustained by intrinsic student interest alone. As with traditional classroom activities, when encouragement does not pull students to participate, then the threat of a low grade will probably push them. Yet even when grades were given for posting messages, about one of five students did not post.

Topics and replies

Message counts tell us nothing about the form or quality of newsgroup discussions. In order to get some idea of their form and quality, we undertook a series of analyses to examine what was said, beginning with an analysis of the form of discussion. An interactive newsgroup discussion, like a good dialogue, would likely generate many replies per topic. In contrast, a newsgroup with many topics and no replies would indicate that students were engaging in monologues, and showing little interest in responding to each other. Keeping in mind the caution noted in Study 1 regarding replies posted as topics by mistake or otherwise, we counted topics and replies by their content rather than their newsgroup title. Though the work was tedious, it was relatively easy. For example, a topic beginning with "George says in Topic 7 that we should boycott the final exam, but I disagree" was coded as a reply. Of the 13% topics we considered replies, and replies we considered topics, we agreed on the classification of all but about 8% of them.

In total, there were 2,453 topics and 4,018 replies posted by students in the 40 courses, an average of 1.64 replies per topic. Instructors (including teaching assistants) posted a total of 129 topics and 36 replies or 0.28 replies per topic. Students, it seems, are more likely to engage in newsgroup dialogues than are their instructors.

Not surprisingly, the number of topics posted by students in the 40 courses was highly correlated with the number of student replies, r (38) = +.99. However, the relationship was weaker for topics and replies generated by instructors, r (38) = +0.55. There was no reliable correlation between the number of topics posted by instructors and the number of student replies, r (38) = - .13, p > .10, nor was there a reliable correlation between the number of topics posted by students and the number of instructor replies, r (38) = +0.03, p > .10. These results indicate that instructors were not reliable participants in course newsgroups, that students and instructors did not engage in much dialogue, and that most newsgroup dialogues (topics + replies) occurred only among students.

How active were these dialogues? As we coded the data we noticed that some topics had many replies and others had none; some replies came within minutes of posting a topic and some took several days; some postings were only a few words and some were several paragraphs. The content of topics and replies varied widely as well. In order to investigate the distribution of replies to each topic, the length of time to reply, the content, and the writing style of the messages, we analyzed a random sample of 628 topics and their 740 replies posted by 262 male and 258 female students as well as 15 male and 12 female instructors. The results are shown in Table 2.

Table 2: Percentage of the 628 Sampled Topics Receiving at Least 1, 2, 3, ... 10 replies

Topics receiving	Percentage of topics
No replies	59%
At least 1 reply	41%
At least 2 replies	24%
At least 3 replies	17%
At least 4 replies	9%
At least 5 replies	8%
At least 6 replies	6%
At least 7 replies	4%
At least 8 replies	2%
At least 9 replies	2%
At least 10 replies	1%

Perhaps the most striking result shown in Table 2 is almost 3 in 5 of the 628 sampled topics received no reply. Dialogues stimulated by the remaining 41% of the topics were relatively short. As calculated from Table 2, 41-24 = 17% of all 628 topics received only one reply, 7% received only two replies, and the percentage receiving further replies continued to drop in a progression similar to Zipf's Law. Male students wrote 46% of the replies, female students wrote 46%, and instructors wrote 6%.

Length of postings

The 628 topic postings averaged 13.1 lines per topic. The average length of the 740 replies was 8.2 lines. In order to determine if there were differences in length of topics, a 2 x 2 ANOVA was performed, with gender and role (male and female students, male and female instructors) as the independent variables and length of topic as the dependent variable. The 27 instructors who posted at least one topic posted longer topics (mean = 24.0 lines per topic) than did students (mean = 10.7 lines per topic), F (1, 624) = 32.12, p < .01). Females posted longer topics (mean = 15.3 lines per topic) than did males (mean = 11.1 lines per topic), F (1, 624) = 8.42, p < .01. There was no significant gender by role interaction, F (1, 624) = 1.3, p = .25.

Length of time to reply

How long did a student or instructor wait to receive a reply when posting a topic on a course newsgroup? There was a large variation in the length of time to reply to the topics ranging from a few minutes to 73 days. Of the 740 replies to the 628 topics, 82% were posted within a day, 13% were posted within a week, and 5% were posted after a week. The results suggest that newsgroup dialogues are rarely sustained longer than about seven days.

Content of the messages

What were students and instructors discussing in their postings? In order to obtain some sense of the content, we read and classified each of the 1,368 postings from the sample summarized in Table 2 according to its dominant theme. Four categories emerged:

1. course content
   Messages posted by students and instructors directly related to the content of the course were classified in this category. Examples include discussions of lecture topics or assigned readings (e.g., What do you think about Milgram's study of obedience the prof. discussed in class today? George, I think the professor wants us to compare the theories of inflation more than to give examples), course outlines and reading assignments, criteria for grading essays, or samples of exam questions;
2. procedural content
   Operational and administrative questions and answers, such as the time or the location of an exam and requests regarding how to do course assignments, were classified as a procedural request (e.g., When is the mid-term exam? Who has the answer to problem #2? Does anyone have notes for lecture 6?);
3. humour
   Any non-serious messages, including jokes and teasing, that were not related to course material were classified as humour (e.g., Our class is too hot! Let's talk about our Prof.'s necktie!);
4. others
   Miscellaneous comments which included anything that could not be included in the other three categories were categorized as other (e.g., "this is a test" message; greetings; book sales).

About half (49%) of all topics and replies concerned the content of the course, 27% procedural content, 8% humour, and 16% others. The results indicate that as replies to a topic accumulated, the proportion of course content increased (e.g., 34% of the first replies, 60% of the fifth replies) while the proportion of procedural content decreased (e.g., 39% of the first replies, 6% of the fifth replies). Twenty-eight percent of the procedural questions received no reply. Humourous replies remained relatively constant (15-21%) as did others (12-20%). These results suggest that procedural discussions are short-lived, and that a discussion can be sustained only if it focuses on course content.

Writing style

While reading the messages, we noticed that they varied in the style of expression. Some messages were quite personal, filled with "I" and "you", words of greeting and encouragement, and self-revelation. Other messages were strictly business, containing only course material with no personal expressions, greetings, or encouragement. About 30% of the messages had a predominantly personal style. About 28% of the messages were entirely course related with no personal expressions. The remaining 42% of messages had a mix of personal expressions and discussions of course material. About 20% of the 176 topics with no personal expression received at least one reply, whereas about 54% of the 452 messages with either mixed or predominantly personal expression in their writing style received at least one reply. This result suggests that a personal style of writing is better for stimulating discussion.

Study 3: Student Survey

The major results of Study 1 and Study 2 suggest that course newsgroups are not widely used for course discussions. Fewer than 25% of the 2,512 1997/98 courses with newsgroups used the medium regularly for course communication. Fewer than 15% of students in theses courses posted messages to course newsgroup discussions, and fewer than half of their postings were related to course content. It is thus safe to say that the new medium of course newsgroups was not widely adopted by Carleton University students or by their instructors. Why? We wanted to know more about students' and instructors' perceptions and opinions of the newsgroups. So we began to learn about them by circulating a questionnaire asking several questions about students' use of various CHAT facilities, in particular e-mail, the Web, and newsgroups.

Respondents

A course newsgroup questionnaire was distributed by instructors of three courses: a second year Introduction to Social Psychology course, a second year Numerical Methods (engineering) course and a forth year Strategies in Management (business) course. Most students took about 15 minutes to answer the questions. Two-hundred and sixty-five undergraduate students from Carleton University voluntarily completed it: 130 males and 134 females (one did not report his/her gender). The average age of respondents was 23.5 years (mode = 21; range = 19 to 55 years). Sixty percent of the participants were aged between 19 and 22 years.

Twenty percent of respondents were psychology majors, 33% were engineering majors, and 16% were business majors. The remaining 21% were from a wide variety of majors such as law, sociology, music, social work, and biology. Three percent of the respondents were in their first year at the university, 47% in their second year, 28% in their third year, 21% in their fourth year, and 1% in graduate school. Fifty-three percent of the students reported they had a job. Those with a job reported working an average of 10.2 hours per week (range = 3 to 60 hours).

Results

Use of CHAT facilities

Of the 265 students, 258 students reported working with a computer, averaging 12.2 hours per week (range = 1 to over 50 hours per week); 83% reported using a computer at home. Twenty-nine students (12 males and 17 females = 11%) reported spending no time on CHAT or any other computer communication facility. Their most common reasons included lack of a computer or Internet skills, lack of interest, and fear of the medium. The remaining 236 = 89% of respondents reported spending, on average, 5.7 of hours per week connected to CHAT.

The vast majority of students found CHAT easy to use. On a scale from 0 (very easy) to 6 (very difficult), CHAT received an average rating = 1.4; 90% of students rated it no more difficult than 3. The result testifies to the success of CHAT developers in making CHAT user friendly, and that it was not a major impediment to CHAT use.

Sixty-five percent of the students had a modem at home. The 153 students with a modem reported spending more time using CHAT facilities (average = 6.5 hours) than the 83 students without a modem (average = 4.0 hours, t (234) = 1.88, p = 0.06). Correlations between (a) age, school year, and number of courses taken with (b) total amount of time spent on all CHAT facilities did not reveal any significant relationships. However, analyses of variance between (a) gender (M/F) and job (Yes/No), and (b) total time using CHAT revealed one significant difference: Students who did not have a job spent on average 7.3 hours per week using all CHAT facilities, whereas students with a job spent on average 4.1 hours per week, F (1, 225) = 4.6, p < .05.

Male students reported spending on average 2.8 hours per week browsing the Web compared to female students who reported spending on average 1.0 hours, F (1, 184) = 5.07, p < .05. Students who had access to CHAT from home reported spending more time exchanging e-mail messages (average = 5.5 hours per week) than those who did not have access from home (average = 2.4 hours; F (1, 230) = 8.54, p < .01). Interestingly, there was no significant difference between male and female students, or between students with or without modem, in their estimated time spent exchanging e-mail messages with friends or in reading/posting on course newsgroups, a result consistent with the possibility that e-mail to friends had highest priority of all CHAT facilities and thus consumed most CHAT time, while course newsgroups had lowest priority and consumed the least (see Table 3).

Students estimated that they spent on average 2.4 hours per week on e-mail exchange, 1.9 hours browsing the World Wide Web (including library catalogues or other online information sources), 0.7 hours on course newsgroups, 0.7 hours on other newsgroups. Thus, only 12% of estimated student time on CHAT was spent on course newsgroups. Eighty percent the 235 students who responded reported never posting a message on their CHAT course newsgroups. The remaining 20% reported posting on average 1.9 messages during the term (range = 1 to 36 postings). Thirty-eight percent of respondents reported never reading a course newsgroup posting. The remaining 62% reported reading an average of 16.3 messages (range = 1 to 100 messages). There were no significant gender or school-year differences among students regarding the number of their reported message postings or readings on course newsgroups. However, those who had a job posted significantly fewer messages (mean = 1.08) than did those without a job (mean = 2.79), F (1, 233) = 5.66, p < .02. These estimates of newsgroup activity may be exaggerated, as the proportion of students reporting that they posted (20%) was somewhat higher than the 13% of students from Study 2 who actually did so. Thus, the proportion of course newsgroup readers may be proportionally lower than the reported 68%. Even so, the result indicates that newsgroups attract about 3.5 times the number of readers as writers. Yet almost two out of five students confessed never reading a single newsgroup posting.

What might account for students not posting on CHAT course newsgroups? Of the 265 respondents, 108 (41%) provided reasons for not posting. Forty-four percent of these 108 respondents indicated lack of motivation related to the paucity of postings by other students, boring discussions, or personal laziness. Twenty-one percent reported reasons of ignorance: they never learned how to use their CHAT account, did not know about the CHAT Course Newsgroups or how to access them, or did not know how to use newsgroup commands. Nineteen percent mentioned organizational reasons. The two most common reasons were: (1) professors did not recommend their newsgroups or do not post topics on them; (2) insufficient time to read or post on course newsgroups. Sixteen percent reported that they preferred alternative means such as personal contact of asking questions or making comments about their courses.

If students were eschewing their course newsgroups for course-related exchanges, were they instead using e-mail? Forty-four percent of the 235 students reported sending e-mail messages to at least one of their professors, posting an average of 4.8 messages (about one per course) to them during the term. About 20% of the students reported sending e-mail messages to at least one of their TAs, averaging 5.9 messages to them during the term. Finally, 70% of the students reported sending e-mail messages to their fellow students during the term, estimating an average of 50.4 messages posted to them. The results indicate that students use e-mail primarily to communicate with each other, and much less to communicate with their course instructors.

Because about one third of these estimates came from students enrolled in the Social Psychology courses taught by the authors, we were able to compare their estimates with our records. We received e-mail from only 8% of the students in that course, and these students averaged only 2.1 postings to us during the term. Almost all of these messages concerned missed exams, clarification of items in the course outline, or requests for special grading considerations. Indeed, of the 163 messages received by about 1,000 students enrolled in four terms of the course, only one message concerned an idea presented in lecture, and only one concerned an idea from the text.

Evaluations of CHAT Facilities

The questionnaire asked students to rate the importance of five CHAT functions on a scale from 0 = not important to 6 = very important. Table 3 shows that the 230 students who answered rated course newsgroups the least important function of CHAT whereas they evaluated the facility of exchanging e-mail messages to friends and access to the Web the two most important. The result parallels the estimated time using e-mail, Web, and newsgroup facilities reported above.

Table 3: Students' Average Ratings of the Importance of CHAT Facilities
(0 = not important, 6 = very important)

	Average rating of importance
Facility	Male (n =119)	Female (n = 111)	All (N = 230)
e-mail access to students & friends	5.2	5.0	5.1
access to Web	5.1	4.4	4.7
e-mail access to professors	4.2	4.5	4.3
e-mail access to teaching assistants	3.7	4.0	3.9
access to course newsgroups	3.4	4.0	3.7

The questionnaire also asked students to rate nine different ways of getting answers related to their course material on a scale from 0 = "not at all helpful" to 6 = "extremely helpful". Of the nine alternatives, the highest rated, on average, was "asking the professor during office hours" (average rating = 4.7), followed by "sending e-mail to professor" (4.6), and "asking professor in class " (4.5). The average rating of "posting a message/question on course newsgroups" (2.6) ranked ninth out of nine alternatives, well below alternatives such as "asking other students in class" (4.0) and "sending e-mail to TAs" (3.7). There were no significant differences among students in their helpfulness ratings of asking questions in class, during office hours, or posting on CHAT course newsgroups according to students' gender, school year, job, or home access to CHAT.

How do students rate course newsgroups as sources of information for education compared to other sources such as lectures, textbooks, and scientific publications? As shown in Table 4, students rated lectures as the most important learning resource whereas they rated CHAT discussions as the least important of all ten alternatives, ranking well below such traditional sources as labs, office hours, and teaching assistants. There were no significant differences according to gender, school year, time, or easy access to CHAT among respondents regarding their average ratings of the importance of lectures, textbooks, and study groups.

Table 4: Average Ratings of the Importance of Learning Resources at the University
(0 = not important, 6 = very important)

	Average Importance Ratings
Learning Resources	Males	Females	Total
lectures	5.2	5.5	5.4
textbooks	5.3	5.3	5.3
students	4.3	4.0	4.1
lab & tutorial	4.0	4.0	4.0
office hours	3.7	4.3	4.0
teaching assistants	3.6	4.0	3.8
study groups	3.3	3.8	3.5
Word Wide Web	3.2	3.4	3.3
journals & publications	2.7	3.4	3.0
CHAT course newsgroups	2.2	2.7	2.5

Do students who read CHAT course newsgroups show a greater tendency to indicate that the newsgroups are helpful for learning than students who read none? Of the 156 students who indicated they read at least one newsgroup posting, 76% reported they believed the newsgroups were helpful for learning. Of the 59 students who reported they did not read any newsgroup posting, only 47% indicated they believed the newsgroups were helpful for learning. The difference is highly significant, Chi Square (1) = 15.6, p < .001, and suggests that students who use course newsgroups find them more helpful for learning than students who do not use them. In a follow-up question, thirteen students noted why it is helpful and 9 noted why it is not. The most common themes for newsgroups being helpful were (1) they help group communication (n = 6), and (2) they are convenient for communication (n = 5). The most common theme for newsgroups not being helpful was that they take too much time (n = 6).

In order to learn what might encourage students to post more often on their course newsgroups, we asked the respondents to rate a series of possible encouragements on a scale from 0 = "not encouraging" to 6 = "very encouraging". By far the most highly rated encouragements were (1) more messages for discussion and (2) feedback. Students rated postings and replies from professors (averages = 4.2 and 4.5 respectively) as significantly more encouraging than postings and replies from teaching assistants (averages = 3.5 and 4.2) and from fellow students (averages = 3.1 and 3.4), F (2, 205) = 44.8, p < .001 for postings and F (2, 205) = 57.5, p < .001 for replies. The results suggest that the professors and teaching assistants can provide more encouragement, and perhaps reward, for students to post course newsgroup messages than can fellow students, and suggests why course newsgroups cannot "run themselves" without instructor involvement. Interestingly, respondents consistently rated replies as more encouraging than postings, suggesting that feedback is more important than leadership. Additional analyses revealed no significant differences according to gender, job, or home access to CHAT on student's ratings of what might encourage them to use course newsgroups for posting their questions and comments.

Study 4: Interviews with Students and Professors

The results of Study 3 reinforce the findings of the first two studies: though students who use them like them, few make use of course newsgroups, and the many who don't use them judge them to be among the least important educational media. In addition, Study 3 gives some insight into why course newsgroups are not popular. The most common reasons were lack of time, especially for working students, and lack of interest in the content. Paradoxically, however, suggestions for encouraging more course newsgroup participation focused on increasing the amount of newsgroup content and feedback to postings, especially feedback from their instructors. The suggestions imply that students are more likely to post if more students post, and if more instructors reply. Posting and replying require time - time taken from other activities. Though respondents noted lack of time for newsgroups, they still managed to spend a lot of it exchanging e-mail with classmates and friends. Why would students exchange messages with individuals as e-mail but not with several individuals in a course newsgroup? Would instructors find the time to reply to more student postings? What else might be done to increase the use of course newsgroups for education? These and similar questions prompted us to pursue one more study, a series of interviews with students and instructors, half of whom were active newsgroup enthusiasts and half of whom were not. We were especially interested in their recommendations for increasing course newsgroup participation rates and the quality of newsgroup discussions.

Methods

Participants

We interviewed 21 students and 20 instructors. Eleven students (5 males, 6 females) were active participants in newsgroup discussions as assessed by authorship of postings examined in Study 1, and ten (5 males, 5 females) never participated. About half of the ten inactive students came from the Social Psychology course we taught; the others were found by word of mouth. Students were in second, third, fourth, and graduate programmes, nine majoring in Psychology, three in Journalism, two in Computer Science, and one each in Sociology, Chemistry, Biology, Anthropology, Political Science, Law, and Engineering. The average age of students was 25 years, and ranged from age 19 to 52.

As with active students, we obtained the names of instructors with active newsgroups by searching the active newsgroups noted in Study 1. Eleven active instructors (5 males, 6 females) agreed to be interviewed. Some of the nine inactive instructors (4 males, 5 females) were found from their barren newsgroups; others were recommended to us by their colleagues as opposed to the use of Internet for education or as self-described technophobics. Six of the twenty professors taught in the Psychology Department, two from each of Social Work, Journalism, and Biology, and one each from Engineering, Communication, Sociology, Linguistics, Law, Philosophy, Music, and Architecture and Design. They had taught for an average of ten years, ranging from two to over thirty years. The average enrollment in their classes was 70 students, and ranged from 20 to 400. Seven of the 11 active instructors maintained one course newsgroup. The remaining four active instructors maintained two.

Interview Questions

The interview questions for the active student contributors to course newsgroups gathered information about five themes:

1. how much time they devote to course newsgroups;
2. perceived differences between face-to-face and newsgroup discussions;
3. the educational benefits they expect from course newsgroups;
4. sources of motivation to contribute to course newsgroups; and,
5. suggestions to improve course newsgroups for education.

The interview questions for inactive participants obtained information about two themes:

1. why they did not participate on course newsgroups; and,
2. what changes might have encouraged them to contribute to course newsgroups.

Similarly, the interview questions for the instructors who had an active course newsgroup gathered information about five themes:

1. how much time they devote to course newsgroups for their teaching;
2. perceived differences between face-to-face and newsgroup discussions;
3. the educational benefits they expect from course newsgroups;
4. how they motivate students to contribute to course newsgroups; and,
5. suggestions to improve course newsgroups for education.

The interview questions for the nine inactive instructors obtained information about the following two themes:

1. why they did not make use of their course newsgroups; and,
2. what changes might have encouraged them to use course newsgroups for their teaching.

Procedure

The active student contributors to course newsgroups were contacted via e-mail or telephone. We explained to them the purpose of the study before asking them to participate. Inactive students were selected in response to a call for participants during an Introductory Social Psychology course lecture, or through word of mouth. Students were interviewed individually in a university office. The active instructors were contacted via e-mail or telephone. We explained to them the purpose of the study and asked them whether they would participate/share their experience of course newsgroup activity with us. We interviewed the agreeable instructors in their office.

All interviews were semi-structured, and were tape recorded for later transcription, coding, and analysis. All interview themes noted above were addressed in each interview, and additional follow-up questions were asked in about 75% of the interviews as the occasion warranted. Each interview took 20 to 65 minutes.

Results

Active students

Amount of Time using Course Newsgroups
Eight of the 11 active students reported they were participating in one course newsgroup during the term in which we interviewed them. All of these eight students were required by their instructor to post, and all reported posting an average of 1.9 newsgroup messages per week. The remaining three students, all in Journalism, reported participating in two course newsgroups. Though none of the instructors for these courses required posting, the three students reported posting an average of 3.0 messages per week.

The 11 active students reported spending on average 3.3 hours per week participating in their course newsgroups, but their estimates varied widely from a low of 15 minutes to a high of 10 hours. The average of 3.3 hours per week is considerably greater than the average 0.7 hours reported by all students in Study 3, and serves as evidence of their abnormal activity.

Perceived Differences between Face-To-Face and Newsgroup Discussions
Because active students had, by definition, more experience than inactive students participating in newsgroup discussions, we were interested to learn how the active students compared newsgroups with the traditional face-to-face interactions that newsgroups are supposed to replace. In response to a question about perceived differences between face-to-face and newsgroup discussions, 10 of the 11 active participants gave lists of the advantages of each.

The most often-noted advantage of newsgroup discussions, addressed by four females and one male, was the opportunity for shy people to ask questions without being embarrassed by the presence of others. These students appreciated the ability to post questions without being observed by others and to avoid being judged "stupid" by the questions they asked. One male student said, "I feel more comfortable with [my course newsgroup] because ... I can post while nobody knows me." Three male students noted that newsgroups give them sufficient time to think before answering a question, and encourage them to write well. One female and one male mentioned that newsgroups allow the convenience of communicating with other students when the time is available rather than arranging to meet at the same place and time. Perhaps because of this, another female added that newsgroups greatly improve the possibilities for distance education.

Two other students also mentioned time saved during a discussion. One male mentioned that he liked newsgroup discussions because, "I do not have to wait to talk." In contrast, one female noted that she liked newsgroup discussions because, "I do not have to listen to everybody in the group."

Interestingly, one female pointed out that students have more control of the discussion in newsgroups than in the class. She noted that instructors tend to direct the discussion in the class whereas on a course newsgroup students are able to direct the discussion because few instructors contribute to it. Thus, a course newsgroup can be a place that allows students more control of the discussion than more traditional face-to-face discussion can.

Given these perceived advantages of newsgroup discussions, should we expect active students to prefer them to face-to-face discussions? Probably not. The active students listed an equal number of advantages to face-to-face discussions. Four females and two males noted that face-to-face discussions reduce the degree of misunderstanding because body language allows them to realize whether a person is being serious or sarcastic, but this is difficult in newsgroup discussions. Two males and two females commented that face-to-face discussions take less time because talking is easier than typing. As one male said, "It takes five minutes to answer a question in face-to-face discussion while it takes about three weeks to answer the same question through newsgroup discussion."

Two females and one male said that because there is no printed record in face-to-face discussions, few students would continue to criticize them after the end of class. In contrast, these three students noted their printed comments in a newsgroup could lead to extended criticisms days or weeks after a posting. One female said face-to-face discussion provides real-time interaction, and offers an opportunity for students to think on their feet and improve discussion skills. Another female noted that the chances of receiving an answer for a question raised in face-to-face discussion are much higher than in online discussions.

All 11 active students agreed that course newsgroups couldn't replace face-to-face discussions. Instead, they all insisted that course newsgroups should be used only as a supplement to face-to-face discussion. None of them suggested that face-to-face discussions should supplement newsgroups, thus indicating that the former have priority over the latter. This reinforces the results of Study 3, which indicated that students prefer asking questions face-to-face (in class, during office hours, or in discussion groups) to asking questions on newsgroups.

Educational Benefits Expected from Course Newsgroups
When asked about the educational benefits of course newsgroups, all 11 active participants noted that course newsgroups allow discussion of diverse topics and the exchange of a variety of opinions about an issue. Four students also considered course newsgroups as a medium to analyze ideas from different perspectives. Five students mentioned that course newsgroups provide useful information about course-related matters such as assignments and exam dates. One female said that newsgroups were a good means to interact with others, and another female claimed that online discussions in the course allowed distant students to evaluate their own academic progress by comparing themselves with other students who participated in course newsgroups.

Sources of Motivation to Contribute to Course Newsgroups
Because eight of the 11 active students were required to post, it is not surprising that six said their major source of motivation to participate in a course newsgroup was to receive course credit. But other reasons were also expressed. Two students stated that their motivation for participating was to communicate with classmates, and to help classmates if they had any course problems. One female said that she read and posted on course newsgroups because her friends did. Another female student confessed that her original motivation for participating was to buy and sell textbooks. And one female student indicated that because she was shy, the faceless discussions motivated her to post questions and comments.

Suggestions to Improve Course Newsgroups
When the 11 active students were asked to offer suggestions for improving educational aspects of course newsgroups, three major themes emerged: (1) give course credit for newsgroup participation; (2) provide better management of the newsgroups; and, (3) offer more computer training. Eight of the 11 students suggested that students should receive marks for their participation in course newsgroup discussion. They argued that allocating marks would increase the motivation for students to participate.

Seven students, four males and three females, suggested that professors and teaching assistants should be more involved in course newsgroups by posting questions and comments, assignments, and quiz samples, and by introducing the medium to students. Eight students, four males and four females, mentioned that students need better training in using the Internet. They noted that many students do not know how to use Internet facilities or even how to log onto the system. Seven of these eight students referred to their own frustrations experienced because of difficulty accessing the Internet outside the campus (an hour of busy signals during peak periods) and frequent system malfunctions. They suggested that more workshops on using the Internet be given and that more phone lines be added for home access.

There were a few other suggestions offered as well. One female suggested the use of "emoticons", to indicate humor, frustration, sarcasm, or other emotions. Another female suggested that part of class lecture time be used for course newsgroup discussions.

Inactive Students

Reasons for Not Participating in Course Newsgroups
The ten inactive students revealed a wide variety of reasons for their lack of participation. Eight of the ten inactive students, three females and five males, confessed to being intimidated by posting in a public medium, disliking computers, being bored by the postings, finding few people posting on course newsgroup discussions, and being unable to post anonymously. The result mirrors the finding of Study 3 that about 60% of students reported social or motivational reasons why they did not post course newsgroup messages.

Seven of the ten students mentioned practical limitations working with computers or with the Internet. Included in their comments were lack of skill in how to use a computer or the Internet, forgetting the location of their course newsgroups, and having difficulty in accessing the Internet from home. Two students, a male and a female, also mentioned that they did not post on course newsgroups because they did not feel confident about their typing and writing skills.

There were logistic and motivational reasons as well. Two females said they did not have time to participate in course newsgroups. One male and one female said they did not participate because the instructor did not require it.

Changes that Might Encourage More Newsgroup Contributions
We asked the inactive students under what conditions they would be willing to participate in a course newsgroup discussion. The most common suggestions were:

• give a mark for posting (noted by 2 females and 1 male);
• make a requirement of the course (1 male and 1 female);
• post more course-related material (2 males);
• post more interesting topics (1 male and 1 female);
• show more student postings (2 females);
• give quick answers to questions (1 male).

Other changes suggested by five students were related to technical issues. These included easier and faster connection to CHAT, providing multimedia in newsgroups that allow students to observe colors, diagrams, sound, and pictures; and, simplifying the printing of course newsgroup articles. Two mentioned that they would participate in course newsgroups if they could post anonymously.

Active instructors

Amount of Time Using Course Newsgroups
The 11 active instructors we interviewed had employed newsgroups for their courses between one and five years. One male and one female instructor were very active using their course newsgroups in previous terms but had stopped or decreased their contributions. One did so because she lacked the necessary time; the other did so because he became discouraged by students' low rate of participation. They both mentioned that before reducing their involvement, they had been spending one or two hours per day, including weekends, to stimulate discussions on their course newsgroups by initiating new topics or commenting on students' postings or otherwise encouraging them to participate in the group discussions.

Active instructors reported checking their course newsgroups from two to 15 hours per week (average = 5.8 hours). Eight instructors mentioned they read their course newsgroups every day, and read all the students' postings. One checked twice a week. The two who had decreased their involvement reported checking their course newsgroups only a few times during the current term. One other instructor mentioned that work on his course newsgroup had become so time consuming that he asked his teaching assistant to help check newsgroup activity and post messages.

All 11 active instructors commented at length about the amount of time and extra workload required maintaining their course newsgroups, especially five males and one female who were also developing Web pages for their courses. Eight instructors noted that they volunteered their time to develop computer technology for their teaching, but feared that their intellectual property (course outlines, lecture notes, etc.) would be used by others without their consent. Five of the eight also feared that their intellectual property would be claimed as university property when posted on the Internet. The 11 active instructors mentioned that they posted occasionally on course newsgroups to announce course news or to answer students' questions they believed they should answer. However, all said they usually let students discuss and answer each other's questions or comment about the course materials, hoping to see students carry the discussion on their own.

Perceived Differences between Face-to-Face and Course Newsgroup Discussions
We asked the 11 active instructors if they believe the course newsgroups would replace face-to-face discussions. Almost all answered by first commenting on the differences they observed between face-to-face and newsgroup discussions, and their answers paralleled those given in the student interviews. According to six instructors, face-to-face discussions accelerate group involvement and enhance teaching more than do discussions in newsgroups. Through body gestures, voice, and tone in face-to-face discussions, students increase their interaction with the instructors and fellow students and this results in a better understanding of course material. However, instructors also reported limitations of face-to-face discussions. Six instructors noted that face-to-face discussions demand coordination of meeting time and place which is often difficult. Five instructors noted that face-to-face discussions could work very successfully for small classes, but are very difficult and sometimes impossible for large classes. In large classes there is not enough time to answer students' questions, and it is not possible to arrange many small discussion groups to suit all students simply because instructors and teaching assistants do not have enough time.

All 11 active instructors believed that course newsgroups can be an effective supplement for discussions in large classes and for communicating with professors without having to meet them in certain places at certain times. They all noted that course newsgroups could provide dissemination of information and communication in a way especially convenient for students far from campus. In this regard, one professor said: "The great advantage of a course newsgroup is that a single mother student can communicate with me from home and does not have to wait at my door to ask her questions."

Instructors also discussed some disadvantages of course newsgroups that limit student involvement in course discussions. According to one professor, a newsgroup's reliance on writing would intimidate students who do not feel confident about their writing ability. Another professor mentioned that course newsgroups would not be advantageous for small classes because in small classes there is often enough time for communication with instructors and among students. Two mentioned lack of visual cues. One said, "A course newsgroup is a faceless conversation based on writing style not talking style" while another reported, "A newsgroup has no body language and social context. It does not allow students to see the differences between joke or sarcasm." One instructor considered it a disadvantage that students had the freedom to criticize the course in public. And all 11 mentioned the amount of time required monitoring newsgroup discussions.

Perhaps because of the complementary strengths and limitations of the two media, eight instructors believed that newsgroups could supplement but not replace face-to-face meetings; only three believed newsgroups could actually replace them. Two of the eight mentioned that newsgroups could replace face-to-face discussions under some conditions: one suggested replacement if the newsgroups displayed pictures of participants; another said they could be a replacement if newsgroups were well-moderated.

Educational Benefits Expected from Course Newsgroups
In order to learn about instructors' expectations of course newsgroups, we asked them what educational benefits they might expect of the medium. Their answers covered several areas of education ranging from communication among students, learning terminology, asking questions about the course materials and sharing similar problems, to ideas such as freedom from class and textbook material.

Five professors mentioned communication aspects of the Internet as an important tool for learning. Their focus on communication, however, varied. For example, a female instructor focused on students sharing personal experiences and learning from each other, whereas a male instructor emphasized communication among students and instructors or TAs at a convenient time for those who are off campus (e.g., students watching course lectures via the local Instructional Television cable channel). One instructor also emphasized communication and discussion of the course among students as an important means to create new topics and new thoughts. Another instructor emphasized communication as a facilitator for group learning through sharing understanding of course materials. According to her, students ask questions and learn from each other through their dialogue. One instructor expected students to learn course materials using the newsgroup while another expected students to learn how to express themselves.

The second theme of expected benefits, mentioned by four instructors, emphasized utilization of communication technology. One instructor suggested that his course is about Internet technology; thus, students are required to learn about computer communication. The other three instructors, however, focused more on the importance of familiarity with technology among female students. They emphasized that can help remove the so-called "technology phobia" among women.

A third theme of instructors' educational expectations focused on learning. One instructor mentioned that course newsgroups allow faster and easier learning of terminology. Another stressed that students can learn about the course through the questions and answers posted on course newsgroups. One instructor also emphasized that course newsgroups allow student learning through sharing their experiences with each other.

A fourth theme of expectations concerned easy access to information. Two professors focused on the accessibility of information. They mainly considered information about course materials, but they also mentioned the usefulness of having access to general information through other newsgroups or information sources on the Internet.

In addition to the expectations above, professors reported a few others. One instructor added that computer communication would provide a democratic approach to discussions of ideas and thoughts and would allocate equal opportunity for each student to speak. She said: "I expected that course newsgroups would give more freedom to talk, provide more democratic, mutual intellectual discussions ... ." Another instructor addressed the issue of saving time and natural resources. He mentioned that course newsgroups could result in having classes with no paper and perhaps no textbook, and would also save time because professors would answer repeated questions once. One instructor also mentioned that course newsgroups would provide her with information for her research interest in situated learning.

The rich variety of expectations might indicate two conclusions: (1) computer communication has great potential, and (2) instructors are optimistic about the potential being fulfilled. However, the interviews with instructors indicated their disappointment with the applicability of current computer and communication technology. All eleven instructors expressed their disappointment, although from different viewpoints, at the failure of newsgroup activity to fulfill their expectations. Here are three typical statements:

"I hoped that CHAT course newsgroup would give students the chance to communicate with me and with other students, and I hoped that this intention results in creating new topics and thoughts, but I never had success with CHAT course newsgroup."

"I expected that CHAT course newsgroup would provide virtual discussions among hundreds of students, especially students taking the course on television, but students are not interested in discussing ideas, they are more interested in gossiping and socializing."

"[Instructional Television] students are usually a week behind the class. The CHAT course newsgroup gives them an idea of the class, enables them to contact TAs. So students can post their questions and TAs would respond, but the problem is that TAs are resistant to communicate with students."

The remaining eight instructors expressed variations of these sentiments. Five expressed frustration with students' lack of involvement in the newsgroup. Three more expressed disappointment with their teaching assistant's lack of involvement in the newsgroup.

Motivating Students to Contribute to Course Newsgroups
Instructors were asked to discuss the ways they have tried to motivate students to participate in their course newsgroups. There were two popular replies: (1) giving students marks for course newsgroup participation, and (2) verbally encouraging them to participate. Six instructors had tried the first, allocating from 5 to 20% of the course mark for student participation. Of these six, two left it as a grading option (optional course credit), whereas the remaining four made it a requirement.

The two instructors who offered optional course credit for newsgroup participation were disappointed that fewer than 30% of their students chose the option, even when the quality of their postings was not assessed, and were perplexed why some students missed an easy opportunity to raise their marks. The four instructors who required students to post in their course newsgroups graded students on weekly or biweekly postings. The four instructors stated that the main purpose for requiring postings was to help students get involved in communication technology and to lower the fear of technology, especially among female students. They mentioned that because evaluation of the quality or content of the postings might intimidate students, they marked students only according to number of postings and not according to the quality of their content. Even so, one of the four instructors said that although posting was an easy way of gaining a few marks, one of her 25 students never posted. Another instructor said that he was disappointed when several students dropped his course when they learned they had to post.

The remaining five of the 11 active instructors had chosen not to give marks but instead to encourage students verbally to participate in their course newsgroups. The encouragement included announcements in class about the existence of the course newsgroup and about its usefulness for learning. These professors noted that they would post quiz samples, student questions asked during office hours or via personal e-mail, and course announcements on their course newsgroups for further encouragement. One professor said that he would mention student names, questions, and comments posted on course newsgroup during lectures in order to encourage other students to post. One professor said that he had given up marking students for their postings because marking is a very difficult task: "I used to give a mark for course newsgroup participation but not any more. It is so difficult to evaluate students' postings: should I grade their quality or quantity? Should I grade the number of postings read?"

All five instructors who gave verbal encouragement for newsgroup participation mentioned that discussions were generally not successful. They noted all the discussions evolved into exchanges among only a small minority of students willing to continue posting. One professor said: "I do encourage my students by writing about myself and by posting questions and comments, but it motivates only a few."

Suggestions to Improve Course Newsgroups for Education
The 11 active instructors' suggestions can be classified into three main categories: (1) workload and protection of intellectual property; (2) organizational; and, (3) technical. Each instructor addressed at least two of these three categories. All 11 instructors mentioned the large amount of work needed to maintain course newsgroups, especially when the instructor has a Web site to maintain for the course as well. The instructors emphasized the many hours of extra work needed to monitor their newsgroups and to develop course Web pages. Eight instructors believed that the course newsgroup workload should be taken into consideration in job evaluations and salary. These eight instructors also mentioned that material they produce for their course newsgroups and Web pages should be their own intellectual property and that they should have control over its distribution and use. They believed the university should provide protection for instructors' intellectual property by ensuring that it is copyrighted in the instructor's name.

The organization of the Internet facilities was the second most common theme from professors to improve course newsgroups. Of the seven instructors who mentioned organizational issues, five suggested that use of course newsgroups should be required for all courses, and that teaching assistants should be responsible to moderate course newsgroup discussions. These instructors suggested that teaching assistants should take more responsibility in stimulating course newsgroup discussions by posting comments and answering questions. Two of these instructors expressed strong dissatisfaction with their teaching assistants who showed no interest in facilitating course newsgroups. However, contrary to the five who suggested that course newsgroups should be a course requirement, one instructor suggested that course newsgroups should not be used for all courses. She said: "course newsgroups should be for those professors who would like to hear students' opinions, discussions, and free writing."

Seven instructors offered more organizational suggestions. One said, "course newsgroup could be a great advantage to deliver courses to students who live in other countries and are interested in taking courses from Carleton. By using course newsgroups to reach students overseas, international students do not have to overcome many hassles to come here for education." Another instructor suggested that course newsgroups should be organized in a way that allows anonymous postings in order to encourage shy students to post. Yet another instructor asked the computer centre not to remove discussions on course newsgroups at the end of the academic year. Finally, one instructor noted: "course newsgroups should not be dominated by males."

The third most common theme of suggestions for improving course newsgroups covered technical issues. Six instructors suggested there should be more phone lines available to dial into the Internet, especially during peak hours at night. After noting that it is now difficult to edit postings on the Internet and impossible to change postings later, three instructors suggested a feature to allow authors to retain, delete, or edit their messages after posting. Two instructors suggested more training opportunities for instructors to improve their use of computer technology for their courses. One of them said: "The University should provide training for professors, especially those with technophobia. We need more training than students do." One instructor also suggested that there should be a feature on the Internet that allows students to login and work on a group project.

In summary, the interviews with active instructors uncover four interesting themes. These instructors have high expectations of newsgroups to enhance learning. However, they are frustrated by the time it takes to maintain their course newsgroups, time they do not have. The instructors are concerned about the ownership and copyright of their course material. Finally, the instructors are discouraged by the lack of student participation in course newsgroups and suggest that the most practical way to encourage more participation is to make it a requirement.

Inactive instructors

Reasons for Not Using Course Newsgroups
Recall that nine university instructors we interviewed, five females and four males, did not maintain a course newsgroup. Their reasons ranged from pedagogical beliefs and organizational issues to career concerns.

Six of the nine inactive instructors stressed the importance of personal contact with students as a valuable teaching technique. They mentioned that only personal contact would generate creative thoughts and scholarly interactions among students. According to these instructors, course newsgroups isolate professors from their students. All six also noted that they did not teach courses on Carleton's Instructional Television channel for a similar reason. One instructor said: "I am not opposed to the Internet or technology but I stick to the traditional notion that the most powerful pedagogy is human interaction, and I do my best to develop my personal contact with my students. There is no substitute for personal contact." Another instructor said: "Only by personal interaction will students learn how to think and how to answer problems." One of the six instructors additionally suggested that the university should encourage more opportunities for personal interaction rather than encouraging more computer-mediated activities.

Five of the nine instructors were concerned about the time needed to monitor and maintain course newsgroups. Interestingly, three of these five instructors were "sessional lecturers" who teach part-time, on contract, and for low pay. Because their salary is no more than $3,500 per course (about the same salary of their TAs), all believe there is no reason to add the extra workload of newsgroup maintenance to their job. None of the 11 active instructors were sessional lecturers.

Four of the instructors simply believed that a newsgroup was irrelevant for their course. One instructor in social sciences said: "A course newsgroup is not sufficient for my course, which demands working on group projects." Another instructor in applied science added: "The textbook and assignments are enough for my students' learning, I do not need a course newsgroup for my courses because it is easier to talk to students and explain the questions."

Three instructors expressed concerns about their career and its security. They mentioned that the extra work to maintain a course newsgroup would not be taken into account in promotion, and instead might result in less control of their own intellectual property. For example, one instructor said: "if class information goes online, then students have no reason to come to lectures."

Some miscellaneous concerns also emerged from the interviews with the inactive instructors. One instructor mentioned that he would not consider using course newsgroups as long as he does not receive complaints from students. Another instructor mentioned that she is afraid to use a course newsgroup because the publicity of names and e-mail addresses might result in receiving many unwanted messages such as commercial advertisements and/or anonymous harassment. Yet, another instructor said: "Students in our department are not interested in using course newsgroups. Thus I do not feel I should use it."

Changes that Might Encourage Contributions to Course Newsgroups
The nine inactive instructors did not address any technical or training issues that might lead them to use course newsgroups. However, they mentioned a variety of other changes or conditions that might encourage them to consider using course newsgroups for their teaching. Three instructors mentioned that they should first think about the utility of this medium and explore the potential of course newsgroups before they consider using it for their courses. However, one instructor said: "Under no conditions will I try it. I do my best to provide time for office hours and personal interaction with individuals and groups of students. I will never consider using a course newsgroup." One instructor mentioned that if the university provided sufficient training for students and motivated them so that they demanded the use of a course newsgroup, then she would consider using one. Another said if university required her to use a course newsgroup, she would do so. The remaining three instructors said that they might consider use of a course newsgroup if they have to teach large classes with hundreds of students and with teaching assistants who would take the responsibility to monitor it. However, none of these instructors showed any enthusiasm to use computer communication as an educational tool for their teaching.

In summary, the inactive instructors generally lack both time and motivation to use course newsgroups. Their comments regarding the requirement of time and effort to maintain course newsgroups and regarding their career concerns reinforce those of the active instructors.

Discussion

"You can lead a horse to water but you can't make it drink." We are reminded of the saying when reviewing the results of the four studies reported above. Most students believed that the course newsgroup facility on CHAT was easy to use. Most students had access to CHAT from university, from home, or both. Yet, despite their easy access, the vast majority of students never contributed to their course newsgroups unless they were required to as part of their course grade, a finding consistent with that of Althaus (1997). Even when they were required to post, about one in five students never did. And the four in five who did post were more likely to contribute postings with scant relation to course content than to contribute relevant topics or replies.

In short, if we judge course newsgroups be the criterion of posting fecundity, success is the exception, failure is the rule. It is possible, of course, to argue that this criterion is too harsh or narrow. Perhaps a high proportion of students were avidly reading the postings on their course newsgroups even though they did not contribute more. Yet our studies show that the high proportion of students who described course newsgroup postings as boring, suggesting that few students would continue to follow such discussions as the course went on. Perhaps students were engaging in dialogue with instructors or with each other using e-mail, thus accomplishing a major purpose of newsgroups in a more private forum. Student-to-student e-mail is private and cannot be collected for analysis. But the authors' collected e-mail messages from students strongly suggest that it is almost never used as an educational medium.

Difficulties accessing CHAT or its newsgroups account for only a small fraction of the reasons students give for not posting course newsgroup messages. The result implies that the effects of improvements to hardware, software, training or interface design on increasing course newsgroup discussion are now or will soon reach a point of diminishing return.

Why, then, do so few students contribute meaningful content to course newsgroups? Students in Studies 2, 3 and 4 provide two primary reasons. First, students have little intrinsic motivation to post messages in course newsgroups. Second, they find little time to do so. These two reasons are, of course, intertwined. If their estimates are valid, most students in our studies spend between five and six hours each week hooked to the Internet. Students with jobs spend less time, those without jobs spend more, presumably because job demands have priority over CHAT temptations. Yet in both cases, most of whatever time is invested in CHAT is spent exchanging e-mail with friends and surfing the Web rather than exchanging newsgroup messages. The priority reflects a preference consistent with our finding (Study 3) that course newsgroups are listed at the bottom of a list of university learning resources. It also suggests that student uses of the Internet are more related to maintaining friendships and finding entertainment than to the dialogical pursuit of truth. It may well be that these social and recreational uses of the Internet are important for student motivation, just as parties and movies were in times past. Yet we suspect that parties and movies continue in addition to Internet mailing and surfing. So time for mailing and surfing must be taken from other activities. One of these activities seems to be course newsgroup posting.

This conclusion begs the practical question: Given the technological and pedagogical potentials of course newsgroups to engage students in educational discussions, how can students be motivated to find time for reading and posting more course newsgroup contributions? Our studies indicate that the encouragement and involvement of instructors in their course newsgroups helps a little, the promise or threat of a grade helps more. But both have their limits. Although it takes relatively little time or effort for instructors to "talk up" the benefits of course newsgroup participation, our studies indicate that such attitude formation messages rarely have a lasting effect. To sustain student posting activities, instructors must frequently monitor the activities, contribute a good share of encouraging responses to the students who post, post new topics or questions for newsgroup discussion, and send messages to the passives in hopes of cajoling them into submissions. These are, for course, the traditional and recommended activities of a good discussion group leader. To do them properly an instructor needs some skill, lots of time and motivation. Yet the proportion of empty course newsgroups found in Study 1, and our interviews with professors, strongly indicate that most professors do not have the time or the motivation. Indeed, the interviews suggest that there is considerable motivation for instructors not to become involved in course newsgroups or course Web sites, partly because it is normally volunteer work which takes time away from activities related to tenure and promotion, partly because of uncertainties about intellectual property and ownership. If instructors are to encourage students to post, then someone must encourage instructors to do so. As Jaffe (1998) reminds us, the encouragement of university professors is traditionally a job for university administrators. If we can generalize from our interviews, it is not a job that administrators are doing well (see also Young (1998) for similar conclusions about administrators at the University of California at Los Angeles).

The shortcut to encouragement is a grade. Our studies indicate that grade offerings for postings have a much stronger effect on course newsgroup participation rates than do encouraging words from instructors. In addition, grading does not itself require an instructor to invest his or her own time in newsgroup discussions. Grading newsgroup postings would thus seem a more cost-effective means of increasing students' posting rates than would social encouragement. But grading does require someone to read and evaluate the postings, and the administrative overhead of tracking postings and grades can greatly reduce the time advantage that grading has over social encouragement. Though our counts are small, they indicate that most instructors offering grades for postings simply count postings and do not attempt the time-consuming chore of evaluating the quality of their content. It would not take students long to find a winning strategy for this quantitative game. The students reported in Study 1 who played the "Who Will Post the Last Posting?" game in their course newsgroup with variations of "I win!" thematic content would, by the standards of posting frequency, receive higher marks for their discussion group than would students choosing not to play, so it would be foolish not to. Indeed, there is probably a shareware program now on the Internet that will automatically post canned messages to course newsgroups on behalf of lazy students, thus boosting their participation rate and grade.

Assuming, therefore, that the quality of content must eventually be graded, someone must read and evaluate it, and to do so, someone must find the time. Alas, a common side effect of grading for quality is an increase in quantity or length, especially from students who believe that the more they post, the greater the chances they will post something of value. This may not be a problem in courses with few students, those intense little seminars of 5-10 participants who expect interaction and are happy to do it from home. It becomes a big problem in populous courses, the thousand-student monsters, when participants are coerced into posting on their newsgroups by the threat of a low grade. How can anyone evaluate the quality of thousands of postings generated by students in a large class? Indeed, how can students be expected to read all this stuff? If professors and their teaching assistants cannot or will not find the time, students might be left to grade each other, thus giving them license to develop their own norms of quality which may not coincide with those of minimal university standards. It seems clear that great care must be taken in setting the conditions under which grades for course newsgroup postings are given.

We believe it is instructive that even the students in our studies who were active in their course newsgroups did not endorse the newsgroups as substitutes for face-to-face discussion. The general agreement among both students and instructors that course newsgroups should be no more than supplemental to traditional instruction methods might be seen as a sign of evolutionary lag; perhaps future generations of students will be happy to endorse the substitution. In the meantime, students and instructors interested in course newsgroups will be faced with two time-consuming tasks: participating in, or grading, both face-to-face discussions and newsgroup discussions. Because time does not expand to fit the tasks available, time spent by students or instructors on at least one of these forums is likely to be insufficient to maintain good discussions in the other.

Students may appreciate the asynchronous conveniences of course newsgroups, but the scheduling problems they solve may be offset by their susceptibility to interruptions. When we are faced with two tasks, one of which can be postponed and one of which cannot, we tend to postpone the former. The asynchronous nature of newsgroup participation allows students to postpone their course newsgroup activities when other demands on their time are more pressing (Thorngate, 1991, 1997). These other demands are almost always pressing, so it is not surprising that newsgroup participation is often put on hold. In contrast, synchronous discussions by definition require participants to meet at the same time and thus, despite the inconvenience of this constraint, force participants to postpone other tasks. The students in our studies who admitted that they were too lazy to participate in asynchronous course newsgroup discussions might be more inclined to participate in synchronous discussions knowing it was their only chance.

Synchronous discussions can, of course, occur on the Internet via chat facilities as well as in the classroom. But talking is 5-10 times faster than typing. The increase in speed often compensates for extra time taken to arrive at a face-to-face discussion, and it usually adds considerable life to the discussion as well. An exciting one-hour class discussion might be quite boring when stretched to a five-hour online chat session, and would be impossible if students did not have five consecutive hours to spare.

In sum, the results of our studies add to the growing corpus of literature suggesting that, as the technological and logistic problems of Internet access are solved, the more difficult social and organizational problems of communication become more evident. The success or failure of the course newsgroup as a forum for pedagogically effective exchanges increasingly rests on social and psychological factors rather than technological ones. In order to develop course newsgroups for effective pedagogical ends, university administrators must find the means to motivate their instructors, instructors must find the means to motivate their students, and everyone must find the time to make course newsgroups work. Until then, course newsgroups will remain time sinks (Whittaker & Sidner, 1997) and their pedagogical advantages will rarely be exploited.

About the Authors

Fatemeh Bagherian received her PhD from the Psychology Department at Carleton University in 1999. She is currently a researcher at the Institute for Research in Planning and Development, Tehran, Iran.
E-mail: fatemeh_b@yahoo.com

Professor Warren Thorngate is a faculty member in the Psychology Department at Carleton University, Ottawa, Ontario, Canada.
E-mail: warrent@ccs.carleton.ca

Note

Data reported in this article were collected by FB as part of her PhD dissertation research, supervised by WT. The report was written by WT.

References

S.L. Althaus, 1997. "Computer-mediated communication in the university classroom: An experiment with on-line discussions," Communication Education, volume 49, pp. 158-174. http://dx.doi.org/10.1080/03634529709379088

J.V. Boettcher (editor), 1994. 101 success stories of information technology in higher education: The Joe Wyatt challenge. EDUCOM: MacGraw-Hill.

E. Brent, 1999. "Computers in the undergraduate classroom," Social Science Computer Review, volume 17, number 2, pp. 162-175. http://dx.doi.org/10.1177/089443939901700203

J. Brown, 2000. "Growing up digital," Change (March/April), pp. 11-20; PDF version at http://www.aahe.org/change/digital.pdf http://dx.doi.org/10.1080/00091380009601719

R.E. Clarke, 1985. "Confounding in educational computing research," Journal of Educational Computing Research, volume 1, number 2, pp. 137-148. http://dx.doi.org/10.2190/HC3L-G6YD-BAK9-EQB5

L. Harasim, 1990. "On-line education: An environment for collaboration and intellectual amplification," In: L. Harasim (editor). Online education: Perspectives on a new environment. New York: Praeger, pp. 39-63.

S. Hiltz, 1990. "Evaluating the virtual classroom," In: L. Harasim (editor). Online education: Perspectives on a new environment. New York: Praeger, pp. 133-169.

D. Jaffee, 1998. "Institutionalized resistance to asynchronous learning networks," JALN, volume 2, number 2, accessed 12 January 1999, at http://www.aln.org/alnweb/journal/vol2_issue2/jaffee.htm

S. Kiesler (editor), 1997. Culture of the Internet. Mahvah, N.J.: Erlbaum.

J.A. Kulik, C.L. Kulik, and P.A. Cohen, 1980. "Effectiveness of computer based college teaching: A meta-analysis of findings," Review of Educational Research,, volume 50, pp. 525-544. http://dx.doi.org/10.3102/00346543050004525

R. Mallett, J. Miller, and J. Stewart, 1993. "Proposal to establish CHAT: Carleton Hotline for Administration and Teaching," Unpublished manuscript at http://chat.carleton.ca/index/original_proposal.html

P. McQuillan, 1994. "Computers and pedagogy: the invisible presence," Journal of Curriculum Studies, volume 26, pp. 631-653. http://dx.doi.org/10.1080/0022027940260604

K. Nakajima, 1994. "Computer writing network project in East Asian languages," 101 success stories of information technology in higher education: The Joe Wyatt challenge. EDUCOM: MacGraw-Hill, pp. 347-353.

D.F. Noble, 1998. "Digital diploma mills: The automation of higher education," First Monday, volume 3, number 1 (January), accessed 12 September 1998, at http://firstmonday.org/issues/issue3_1/noble/ http://dx.doi.org/10.5210/fm.v3i1.569

D.N. Perkins, J.L. Schwartz. M.M. West. and M.S. Wiske (editors), 1995. Software goes to school. New York: Oxford University Press.

J. Schofield, 1994. "Barriers to computer usage in secondary school," In: C. Huff and T. Finholt (editors). Social issues in computing: Putting computing in its place. New York: McGraw-Hill, pp. 547-578.

W. Thorngate, 1988. "On paying attention," In: W. Baker, L. Mos, H. Rappard, and H. Stam (editors). Recent trends in theoretical psychology. New York: Springer-Verlag, pp. 247-264.

W. Thorngate, 1990a. "The economy of attention and the development of psychology," Canadian Psychology,, volume 21, pp. 62-70.

W. Thorngate, 1990b. "Got a minute? Attentional limits revisited," Canadian Psychology, volume 21, pp. 97-101.

W. Thorngate, 1991. "Patterns of computer communication: A report on Brucella Network activity," In: J. Frank (editor). Networking in Brucellosis Research. Tokyo: United Nations Univesity Press.

W. Thorngate, 1997. "More than we can know: The attentional economics of Internet use," In: S. Kiesler (editor). Culture of the Internet. Mawah, N.J.: Erlbaum, pp. 296-302.

W. Thorngate and L. Plouffe, 1987. "The consumption of psychological knowledge," In: H. Stam, T. Rogers, and K. Gergen (editors). Metapsychology: The analysis of psychological theory. New York: Hemisphere Press, pp. 61-92.

S. Whittaker and C. Sidner, 1997. "Email overload: exploring personal information management of email" In: S. Kiesler (editor). Culture of the Internet. Mawah, N.J.: Erlbaum, pp. 277-296.

J.R. Young, 1998. "A year of Web pages for every course: UCLA debates their value," Chronicle of Higher Education, volume 44, number 36 (15 May), accessed 12 September 1998, at http://chronicle.com/free/v44/i36/36a02901.htm

---

Editorial history

Paper received 18 June 2000; accepted 4 July 2000.

Copyright ©2000, First Monday

Horses to Water: Why Course Newsgroups Fail by Fatemeh Bagherian and Warren Thorngate
First Monday, volume 5, number 8 (August 2000),
URL: http://firstmonday.org/issues/issue5_8/thorngate/index.html