Listservs in the college science classroom: evaluating participation and "richness" in computer-mediated discourse.

How do instructors motivate students to participate in computer-mediated discussion? If they do participate, how can the quality of their interactions be assessed? This study speaks to these questions by examining online participation and discourse in a science course for preservice teachers. The instructor of an introductory entomology course for preservice teachers implemented online discussions by way of a listserv that was designed to provide students with greater access to important information outside of class. Data was collected from focus groups, written questionnaires, interviews with the instructor, and 182 public listserv messages. Initial student participation was encouraged by the instructor, but participation was modest. The posting of the first mandatory assignment halfway through the course, however, corresponded to a burst period of student activity, yielding a four fold increase in the number of messages authored by students. There was also a seven fold increase in the proportion of discussions that involved at least two student participants and a 50% increase in the proportion of outside references cited within the body of students' messages. This latter finding reflected improvement in the quality of online discourse among students. This evidence suggests that instructors who are interested in listserv participation should make some of their listserv activities mandatory.

**********

Over the last decade, learning networks have been incorporated into classrooms throughout the country. Why? Over two-thirds of 550 educators surveyed reported that "integrating telecommunications into their teaching made a significant and positive difference in how they teach" (Honey & Henriquez, 1993), to the point where, some studies (Hiltz, 1994; Harasim, 1987b; Hsi & Hoadley, 1997) have reported that online environments facilitated learning outcomes that were usually equal or superior to those generated in face to face situations. Part of the reason for this outcome may have been because computer-mediated communication appears to increase motivation (Harasim, 1987a). Students have also reported that computer mediated communication has "allowed for a more democratic environment, increased interaction, and better access to group knowledge and support" (Harasim, 1987a). Despite these reported benefits to students, some students will still not join a listserv (Althaus, 1997). The issues motivating this study are: how do instructors motivate students to participate in computer-mediated discussion? If they do participate, how can the quality of their interaction be assessed by instructors?

In the Fall semester of 1997, an instructor of a college level entomology course for preservice teachers attempted several innovations. One of the innovations was the addition of an online discussion group to a shortened class period. On the first day of class, the instructor informed students that they would be asked to participate on a listserv. The instructor's expectation was that the listserv would enhance discussion about insects as ideal specimens to study within a science classroom and offer a forum for students to ask questions that were not addressed in the shortened class period.

Early on in the course, several of the posted messages were read. After analyzing this early sample of messages on the listserv, three focus questions emerged to guide the rest of this study:

1. Did the level of student participation change over the course of the listserv?

2. If student participation grew, what factor(s) facilitated this growth?

3. Did the quality of student discourse change during the course of the listserv?

The purpose of this study was to track student participation, uncover motivating factors influencing growth in participation, and if possible, find new variables to assess the quality of student discourse on the listserv. The second purpose of the study was to generate a viable hypothesis to explain why the level of student participation changed and draw possible implications for motivating participation.

THEORETICAL PERSPECTIVES

It was only in the early 1980's that colleges began to experiment with computer conferencing for undergraduate course delivery, and by this time, researchers had already completed an extensive review of computer-mediated communication (CMC) and its utility for teachers. In a substantive, long-term review of listservs, Harasim (1990) suggested that several characteristics were unique to this mode of communication. One of these characteristics was a difference in the level of participation observed on a listserv compared with the level of participation found in a traditional classroom setting. The difference in student participation was substantiated again in another survey of 240 teachers and learners on the Internet.

  Of 176 respondents to the question on whether computer mediated
  communication (CMC) is different from the traditional classroom, 90%
  noted that students become active participants; discussions become
  more detailed and deeper; learner-learner group interactions are
  significantly increased, and personal communication among
  participants is increased. (Harasim & Yung, 1993)

According to this survey, the level of student participation in class discussion appeared to improve when there were also opportunities for further discussion through electronic discussion groups outside of the class. This study tracks the amount of participation on a listserv over a semester and notes any changes in the level of student participation over this time.

McComb (1994) also noted that the level of student participation improved (in a class with a listserv component), because the discussion seemed to build on itself and participation was spread among many instead of a few. Walther (1992) proposed that participation is spread among many in electronic discussions because "the lack of social context cues is also conducive to equalized participation. When such cues are absent, actors become disinhibited who would otherwise defer speaking turns to higher-status participants." But Althaus (1997) cautioned that despite the promise of equalized participation, "When [students are] left to choose for themselves, only a small proportion can be expected to participate." Some instructors have attempted to overcome this reluctance to online participation by making listserv assignments mandatory instead of optional; however, Quinn, Mehan, Levin, & Black (1983) suggested that when online participation is mandatory, students may respond with anxiety and resistance. This study adds to our understanding of the nature of mandatory assignments by investigating the impact of mandatory assignments on participation and the nature of discourse on a listserv.

Compared with electronic communication, many would agree that face to face communication is the "richest" form of communication because of the body language used to convey subtle messages and, "the availability of immediate feedback" (Walther, 1992). These attributes of face to face communication make it possible for face to face communicators to converge upon a common understanding that other forms of media cannot accomplish so efficiently. "Richness" is an ambiguous term to describe the quality of communication; however previous work in a different field--communication theory and organizational management, has attempted to describe a similar concept of "information richness" (Daft & Lengel, 1984; Trevino, Lengel, & Daft, 1987). According to these authors, richness is defined as the potential information-carrying capacity of data. Daft and Lengel (1984) offered this example: a wink may be considered a more rich form of communication than a statistical theorem if the wink provides substantial new understanding that the theorem did not. Given this definition and example of information richness, face-face communication receives the highest rating in Daft and Lengel's hierarchy of media richness, compared with written notes and memos (a moderate rating) and computer output (the lowest rating). With the rise in popularity of e-mail, Steinfield and Fulk (1986) proposed that electronic media should be placed above written notes and memos (moderate). The justification for this placement was their belief that electronic media provides rapid feedback and uses natural language but lacks the multiple cues of face to face communication. Researchers may agree (Foulger, 1990) that electronic media can enrich communication in a learning context, especially since the teacher may find it difficult to provide personal attention to every student during the class like they can on a listserv. Given these time constraints, teachers and students using a listserv have to convey their message more through electronic text. Since body language cannot support their meaning, the text must accomplish this, making electronic messages more rich than a low rating would suggest. This study, however, does not attempt to align itself with any particular "camp" on the richness hierarchy; rather, for the purposes of educational work, a broader definition that includes the variety of content and resources conveyed in electronic text and the variety of interaction stimulated by the text is favored.

This study proposes a new variable that could describe one facet of rich communication through the text of the an electronic message and that is the number of references cited in the body of the electronic message. Although the number of reference citations does not capture all that one may intuit by "rich" communication, it is a codable variable that could be used to assess one aspect of the quality of student interaction and discourse. For this study, reference citations in electronic text included references to websites, documentaries, poetry, community events, and school programs to name a few. Telling other students where to access further information about insects shows that students are attempting to connect what they have learned in the classroom to outside resources. It may lead other students to seek out additional references and stimulate further learning. In this way, providing references enriches the message and broadens the scope of available material for students of the course to review. This study analyzes students' messages for reference citations as the listserv progressed.

METHODS

Using Insects in the Classroom is an introductory entomology course offered at a major university in New England. In a Fall semester of this course, the instructors implemented several innovations at once: a distance learning component involving a second remote classroom connected by a videoconferencing link; student independent viewing of instructional video tapes, and a listserv.

The class participants (n=38) ranged from elementary through to secondary education students. The majority of students enrolled in this class, however, were secondary education students who were preservice and inservice teachers. Although entomology may seem at first glance to be an obscure topic for elementary and secondary teachers, the fascination children hold for animals, combined with the comparative ease with which teachers can gain inexpensive access to supplies of insects, makes it an excellent science area for teachers to explore.

The instructor's description of earlier versions of the course from the syllabus stated that:

  This was a traditionally structured course which met one evening a
  week for two and a half hours ... This course was taught in a semi-
  traditional manner using the lecture method as the major form of
  information transfer. In the last few years, an effort was made to
  include more hands-on science.

In designing this year's innovations, the instructor's focus was on students' ownership of their learning--he set a goal that 2/3 of the time be allocated for independent learning. They reduced the scheduled face to face class time from 2.5 to 1.5 hours and increased the expected workload of independent learning outside of the class by the same increment. Students were expected to spend a portion of the extra hour outside class watching instructional videos and another portion on the listserv.

Research Design

The case study was documented by participant observation since the author was enrolled as a student in the course. Participant observation is a research method in which researchers observe behavior in real life settings in which they are participants (Adler & Adler, 1994). The entire course was observed for innovative strategies in college science, and from this case study emerged a descriptive and analytic examination of the listserv. Data was gathered from several sources: the course listserv, in classroom observations, student focus groups, student questionnaires, and instructor interviews. The amount of data collected was: 182 public listserv messages, classroom observation notes of each lesson during a semester, 3 student questionnaires, informal interviews with students, taped transcripts of focus groups (with n=8 students total), two interviews with the instructors, and video taped documentation of two lessons. Triangulated support for findings was sought from these multiple data sources.

Messages were collected from September 16 to December 18; a period of one semester. More than 300 messages were posted in this time frame, but the messages included in this study were restricted to only those that were made public at some point. Private messages were not included in the study, but if any part of the message became public, the public portions were included in all analyses and dated at the time they became public. One hundred and eighty two listserv messages met these criteria.

The 182 public messages were chronologically ordered and then grouped into several emergent categories; five categories are reported here. First, the online interaction cited was categorized as two separate messages even though they were part of one discussion about the same topic. Second, messages were categorized as authored by a "Student" (S) or "Instructor" (I) so that comparisons could be made between the two. For example, the first message was from a student (S) and the response was from the instructor (I):

  (S) I read recently in the Oct/Nov Audubon magazine that scientist
  have introduced a non-indigenous insect to help with the purple
  loose strife problem. Do you have any personal feelings about this
  process? The article said that the success rate was somewhere in
  the 80%-90% range of keeping loose strife from spreading.

  (I) Thanks for this great message. A couple of years ago there was
  a ... student in my class from the Berkshires. She was very much
  involved in using a beetle to control a millfoil plant that has
  become rampant in the lakes of western Mass ... [Her USDA] job is to
  travel throughout Europe and to find insects that attack major weed
  or plant problems we have in the US. Since many of these insects are
  host specific they work quite well. I do believe in using them.

This interaction would be coded as two separate messages with the student being the author of the first message and the instructor being the author of the second message.

Third, codes indicating the "direction of communication" were established, with the first message being identified as student to instructor (S[right arrow]I). Similarly, other relevant codes included I[right arrow]S, S[right arrow]S and I[right arrow]C, where C represented a broad bulletin for the entire class to read. Messages were categorized according to these codes and sub-coded as containing reference citations if students' cited a source of information outside of the class, such as the first message's reference to the Oct/Nov issue of Audubon magazine.

Fourth, a distinction between the first half of the course (Period 1) and the second half (Period 2) was made (Figure 1). This distinction was made because events along the timeline serve as important indicators in the descriptive analysis of the listserv.

[FIGURE 1 OMITTED]

1. The point that separated the first period of the course with the second period of the course was determined to be November 2nd since this date marked the end of the 7th week out of 14 possible weeks online. Thus, each period consisted of 47 days online.

2. All of the public messages sent in the first half of the course, referred to as Period 1 = Sept 16-Nov 01, were compared with all of the messages sent in the second half of the course, Period 2 = Nov 02-Dec 18.

3. The last week of Period 1 also marked the time when a significant event, a mandatory assignment, was introduced. Up until this assignment, students were simply asked to participate in electronic discussions.

The fifth code was a new variable that could describe one facet of rich communication through the text of an electronic message and that was the number of references cited in the body of the electronic message. For the listserv messages in the present study, a reference was coded as a citation to a website, multimedia or televised documentary, poetry, literature, community events and school programs within the text of the electronic message. For example, the following electronic message contains a reference to an Ecological Database of the World's Insect Pathogens available on the Web, and so it would be coded as containing a reference citation.

  No 24 (S) Ecological Database of the World's Insect Pathogens
  http://insectweb.inhs.uiuc.edu/Pathogens/EDWIP/index.html.
  Improvements in global transportation along with increased movement
  of agricultural and horticultural products has allowed more species
  of insect pests to spread around the world.... This site is a
  valuable source of information for scientists involved in microbial
  control research as well as those interested in the possible risks
  of releasing microbial control agents.

After the classroom observation and listserv data was collected and coded, graphs were constructed to represent any trends. Rival hypotheses were eliminated based on their lack of empirical consistency until a single viable hypothesis appeared to explain the trends we observed between Period 1 and Period 2.

RESULTS AND DISCUSSION

When students begin to convey messages to each other through a computer, the parameters governing the nature of discourse are altered. Even though electronic communication cannot convey the subtle cues of body language or the inflections of speech, the first section presented supports previously cited benefits of electronic discussions.

In the first section, the listserv data was examined to report on the length of electronic discussions, time for student reading, reflection, and composition on listservs, and bidding for teacher attention phenomenon. The analysis of the listserv discussion was compared with an analysis of classroom discussions. In the second section, the listserv data was examined to report on changes over time. Specifically, the second section addressed the questions: Did the level of student participation on the listserv change over the course and did the quality of discussion change? Hypotheses were generated in an attempt to explain changes observed on the listserv.

Section 1. Length of electronic discussions, time to compose listserv messages, and opportunities to interact with the instructor.

Length of electronic discussions compared with verbal in class discussions. This analysis compares the length of student discussions from the listserv with the length of student discussions during whole classroom discussion. The length of student discussion from the listserv was tracked by determining the average number of words written per message by students, and in the classroom context, the length of student discussion was determined by the average number of words spoken by students in a typical 90 minute classroom period. A random sampling of listserv messages from the pool of student messages was used to determine the average length of the messages on the listserv. Copied messages meant as a "thread" to the reply; pathway information at the beginning of each message; or names in the closure, were not included as lines of text. From this random sampling, it was determined that on average, electronic messages on the listserv had approximately 18 lines of text and an average of 12 words per line. The average written message on the listserv, according to this calculation, was about 216 words in length.

The average amount of time allotted for whole class discussion was measured to be 22 minutes per 90 minute lesson. In that 22 minutes, an average of 19 questions were asked by students or answered by students. The majority of students' talk in these instances were detected to be less than 4 words in length which would equate to less than 80 words contributed by students during whole class discussion. The average word contribution on the listserv tended to be considerably more than the average student response during whole class discussion, an observation that supports previous findings (Quinn et al., 1983).

More time for reading, reflection, and composition on listservs. The Quinn et al. (1983), study on electronic discussion groups revealed that in one of their typical online sessions, students spent 19 minutes on average just reading messages left by others and an additional 28 minutes composing their own--a total of 47 minutes on average for reading, reflecting, and composing responses on listservs.

Two students from this course reported that listservs also provided them with additional time for communication. For example, one student revealed in the student questionnaire that the listserv appeared to bridge the gap between class meetings: "I felt it was a good way to communicate--with the class + prof--once a week is not that often to meet." A student from the focus group (n=8) also said that,

  I was complaining about the fact that we didn't have enough time to
  interact in the class, and it was the only way I though I could
  actually communicate with students on a regular basis. For me, where
  some people learn very effectively through writing or reading, and
  others through making models, I learn very effectively through
  discussions, and it was an opportunity to have a discussion spread
  out over weeks at times.

It appears that the time available for reading, reflection, and composition on the listserv was welcome by some students from the course who felt that 22 minute whole classroom discussion did not afford enough opportunity for students to interact with their teacher and with other students.

Eliminates bidding. In traditional classroom discussions, the wait time after a question has been posed by an instructor is typically 2-30 seconds before an answer is provided. A 2 to 30 second wait time is not enough time for many students to formulate a thoughtful response to a teacher's questions. Short wait times tend to reward those students who are quick at devising verbal responses. Even students with their hands up "bid" for their teacher's attention to respond, until one student, or a group, is called upon to share their answer first. According to Quinn et al. (1983), electronic discussions eliminate this kind of "bidding" for the teacher's attention. Bidding is potentially eliminated in electronic discussions because there may be more time for every student to formulate and post a response to a teacher question than there is in a class where discussion is synchronous and wait time between the teacher question and student response is 2-30 seconds. When online feedback is immediate, students do not have to bid for the teacher since they can receive individual attention by instructors on e-mail. Quick responses by an instructor on a listserv prevent what Romiszowski and de Hass (1989) found happens when feedback on a listserv is not immediate: procrastination by students to respond or their eventual withdrawal from the discussion group. In the context of the listserv that was examined in the present study, however, the instructor responded to close to 100% of all 182 messages within 48 hours. In this respect, the listserv may have come close to eliminating the bidding phenomena because students had greater than 30 seconds to respond to a teacher's question on the listserv and the teacher was able to provide electronic feedback to every students' posts.

Based on the analysis in the first section, it appears that several parameters governing student discourse may have changed when the learning environment moved out and beyond the classroom walls. In the case of this course, the listserv had longer written messages compared with the length of verbal responses measured in the class. For some students, the electronic medium also appeared to offer more time for reading, reflection, and composition and may have come close to eliminating the "bidding phenomena" for the instructor's attention.

Section 2. Results on level of student participation and quality of online discussion over time.

Focus questions. This section will discuss the three questions that were highlighted at the beginning of this study: Did the level of student participation grow over the course of the listserv? What factor(s) facilitated the growth of the listserv? Did the quality of discussion change?

1. Did the level of student participation change over the course of the listserv?

The data revealed that student participation grew from several students to 29 students by the end of the semester, and the number of student messages per week also increased by the end of the semester as shown in Figure 2. Figure 2 does not show student messages from October 12-18 because there was a holiday during that time period.

Period 1 (from September 16-November 01) had an average of three student messages per week compared with the second half of the course, Period 2 (November 02-December 18), which had an average of 14 student messages per week. This change in the rate of responses represents a 367% proportional increase in the level of student participation from Period 1 to Period 2 on the listserv.

[FIGURE 2 OMITTED]

2. What factor(s) may have facilitated the growth in student participation on the listserv?

This increase in student participation on the listserv in Period 2 may have occurred because a mandatory assignment was given to students on October 26, the last week of Period 1. The assignment was a mandatory online exercise where students were expected to share an experience about reflective teaching. The purpose of sharing personal experiences in education was to help future and practicing teachers learn exemplary teaching practices from one another. The mandatory assignment was introduced to students during the week of October 26, and because the increase in student messages happened immediately after this week, the mandatory reflection assignment may have been an important factor that facilitated the growth of the listserv.

Development of working hypotheses. Consequently, one of the working hypotheses generated in this study was that the mandatory reflection assignment was responsible for the increase in the average number of student messages per week because it was mandatory. Two alternative hypotheses to explain the increase in the average number of student messages per week were that more students had gained access to modems in this time period, causing greater participation and the average number of student messages per week to increase, or, as time passed, students became more comfortable with the idea of posting or more experienced with posting, causing the average number of student messages per week to increase.

The first alternative hypothesis was less likely since in a focus group, students reported that access to modems was not an issue for them because all students were given access to the university's computers at the beginning of the semester. If the second alternative explanation was true, one would expect that as time passed and students gained experience over time, the bars representing student messages in Figure 2 would gradually rise from Period 1 to Period 2. But the shape of Figure 2 does not support this gradual trend. In Period 1, student messages were being posted on average, 3 messages per week, and then suddenly, there was a burst of activity where 21 student messages were posted by a variety of students. Two weeks following the introduction of the mandatory assignment, an average of 17.5 student messages were still being posted.

Does the increase simply reflect required responses to the assignment itself? Even with the students' responses to the mandatory assignment itself subtracted out from Figure 2, the sudden increase in messages was sustained throughout the remainder of Period 2, where a 224% proportional increase was evident (Figure 3).

Other contributing factors which may have amplified this growth may have included, in part, a repeated request by the instructor to post group questions from an activity conducted on the first day of class and generally greater experience on line as time passed. To some limited degree, all of these other factors may have worked in concert to amplify the responses in Period 2. The timing of the burst pattern and sustained increase shown in Figure 3 suggests however, that the principle factor

facilitating the growth of the listserv was the mandatory assignment.

[FIGURE 3 OMITTED]

What makes the mandatory assignment even more noteworthy, is that for some students, it may have motivated them enough to overcome an "activation barrier" and join the listserv as permanent members. Twenty eight percent of all posters began their first post with the mandatory reflection assignment, and of these "slow starters," 50% went on to post at least one more message. In fact, only a quarter of all student messages posted in Period 2 were the assignment itself. Thus, the mandatory reflection assignment may have acted in some way as a "catalyst" to motivate some slow starting students used to posting messages. According to Harasim, Hiltz, Teles, and Turoff (1995), the most important characteristic for student success in this medium is motivation. One factor that appeared to motivate students here was the mandatory assignment.

The following three observations formulate starting points for a plausible hypothesis to explain the greater than 360% increase observed in student participation on the listserv.

Observation 1. More student participants were observed after the mandatory assignment.

Observation 2. More student messages/week were observed after the mandatory reflection assignment.

Observation 3. Some students overcome an "activation barrier" after their response to the mandatory assignment.

The explanation that appears to have the greatest likelihood is that the mandatory assignment was the principal factor in facilitating the growth of participation in the listserv.

3. Did the quality of discussion change?

Modes of interaction. The average ratio of teacher to student talk is about 3:1 in a traditional classroom environment; whereas the ratio of teacher to student contributions on a listserv can be as low as 3:7 (Piburn & Middleton, 1997) or less than 10% (Harasim, 1987b). Observations of verbal interaction present in the classroom were often student to instructor, instructor to student, or instructor to the whole class. These three modes of interaction were examples of discussions that had two participants, where one participant was always an instructor. This kind of interaction with one instructor (the one to many mode), remained the dominant mode of interaction in the classroom discussions that were observed. Not surprisingly, this mode of discussion was mirrored again on the listserv. If a student received a response from anyone on the listserv, 97% of the time in Period 1, the respondent was an instructor. In fact, only two messages in this period were initiated by a student and responded to by another student during this period.

But the nature of interaction among students changed in Period 2. By the end of Period 2, 17 messages were initiated by a student and involved at least one other student respondent. This back and forth messaging between two or more students with regards to the same original message was coded as a "multi-participant" discussion. For example, the three student authors below collectively mounted a response from four other students within two days of their posts:

  I have a colony of the milkweed bugs from class at home and have
  been checking and watching them a bit everyday. A few mornings ago,
  one of the immature bugs turned a bright orange/red ... even his
  legs! I wondered if this was a precursor to molting? Has anyone else
  had this experience ...?

  There have been some very interesting posts on the Monarch Watch
  newsgroup (and others) stating that several Monarch larva have
  recently been found in Texas and California. This is unusual in that
  they should be in diapause-not reproducing. [T]his has been linked
  to El Nino ... what affect might El Nino temperatures have on other
  insect reproduction?... Could it be affecting insect reproduction
  or migration? Will the Monarch's multi-generational migration be in
  jeopardy if they begin to reproduce in Mexico rather than "rest"? My
  best guess is that if the Monarch's in Mexico (and California)
  begin to reproduce, the adults would then die and the larva wouldn't
  have enough food down there?

  Just a quick additional reflection, with an entomological metaphor:
  The centipede was happy quite
  Until a toad in fun said,
  Pray, which leg after which?
  That worked her mind to such a pitch,
  She lay distracted in a ditch
  Considering how to run! (Anonymous)
  The world of details wets an artist's appetite and a scientist's
  curiosity and vice versa.... I once thought that there was a danger
  in focusing too intently on the Parts.
  Upon reflection I now believe the real danger lies in not returning
  after minute inspections to contemplate the Whole.

Table 1 reports the changes in the proportion of multi-participant student discussions from Period 1 to Period 2.

The results indicate an over 650% increase in the proportion of multi-participant student discussion between Period 1 and Period 2.

Quality of discourse. In a self-reported questionnaire following a semester of electronic discussion, the majority of 96 students surveyed from eight courses revealed that the addition of electronic discourse made it more likely that they developed a positive relationship with peers and the instructor, asked teachers and peers a question, and thought before answering (Karayan & Crowe, 1997). Thinking before answering may have led to the next observation reported: an increase in the number of student messages containing references in the text of the message was observed. Some of students' messages contained references to videos, television programs, local /national school programs, magazines, web sites, community events, folklore, and poetry. The following list highlights several messages that contained references:

  No 19 When Susan Mitchell and her students (on Video #10) said that
  the female withholds sperm from some eggs and these eggs become
  male. Why?

  De 15 ... I read recently in the Oct/Nov Audubon magazine that
  scientists have introduced a non indigenous insect to help with the
  purple loose strife problem. Do you have any personal feelings about
  this process? The article said that the success rate was somewhere
  in the 80%-90% range of keeping loose strife from spreading.

  No 24 Ecological Database of the World's Insect Pathogens
  http://insectweb.inhs.uiuc.edu/Pathogens/EDWIP/index.html

  Improvements in global transportation along with increased movement
  of agricultural and horticultural products has allowed more species
  of insect pests to spread around the world.... This site is a
  valuable source of information for scientists involved in microbial
  control research as well as those interested in the possible risks
  of releasing microbial control agents.

  No 26 Singaporeans have been flocking to a nearby Malaysian town to
  feast on a fashionable but expensive snack: termites..... Reader's
  Digest.

This study proposes a new variable that could describe one facet of rich communication through the text of an electronic message and that is the number of references cited in the body of the electronic message. Although the number of reference citations does not capture all that one may intuit by "rich" communication, it is a codable variable that could be used to assess one aspect of the quality of student interaction and discourse. For the listserv messages in the present study, reference citations included websites, documentaries, poetry, community events, and school programs to name a few. Telling other students where to access further information shows that students are attempting to connect what they have learned in the classroom to outside resources. It may lead other students to seek out references and stimulate further learning. In this way, providing references enriches the message and broadens the scope of available material for students in the course to review. In Period 1, 9/19 student messages contained such references. By the end of Period 2, 51/73 student messages contained references. This change represents a 47% increase in the proportion of messages containing reference citations from Period 1 to Period 2.

The evidence suggests that the nature and quality of student discourse appeared to change in two ways in Period 2. The proportion of multi-participant student discussions increased by 667% and the proportion of references cited in the body of student messages increased by 47%. The increase in the proportion of references among student messages seemed to enrich the quality of the discussions, as it made new connections from outside sources to material in the course and consequently, expanded the resource base on insects for everyone. Computer-mediated discourse is normally described as being lean (Trevino et al., 1987), compared with face to face discussion, because it has a lower potential to carry meaningful information. In an attempt to broaden the definition for educational purposes, this study contributes reference citations as an aspect of richness in computer-mediated discourse.

Instructor's electronic messages change. As the proportion of student participants and student messages/week increased on the listserv in Period 2 compared with Period 1, the instructor's messages also changed. In Period 1, 54% of the instructor's messages were addressed to a particular student, but by the end of Period 2, 90% of instructor messages were addressed to a particular student. By addressing the message to a particular student, it may appear to be more personal to the student compared to messages that are not addressed to a particular student but rather broadcast to the entire class.

Also, it is apparent from an examination of the text of these messages that the instructor's message was in response to a student message as opposed to being initiated by the instructor. Figure 4 is consistent with this qualitative observation because the instructor bars roughly follow the same pattern as the student bars in Period 2; whereas in Period 1, the instructor and student messages were incongruent--suggesting that discourse was not likely responsive. An examination of the text in Period 1 confirms this hypothesis.

[FIGURE 4 OMITTED]

The instructor reported that his messages in Period 1 were mostly informative. He stated in an interview that in the first half of the course, he was trying to initiate discussion by disseminating factual information. In Period 2, on the other hand, the instructor's messages appeared to take on a supportive tone and were more like offering opinions than disseminating information. When asked to respond to this observation, the instructor agreed that in Period 2 he was scaffolding student discussion that was already happening. The instructor stated that,

  ... in the second half [of the course], students also had a greater
  knowledge base and were better prepared to pose hypotheses ... Also,
  I was less willing to just give an informational response that
  included factual information. I was hoping that students at this
  stage of the course would answer one anothers' questions.

This change in instructor roles from an "initiator" of discussion to a "respondent" supports the notion of a listserv whose discussion was becoming increasingly student-centered, a goal of the instructor. In a traditional, teacher-centered learning environment, the conversational pattern can follow an initiate-respond-evaluate (IRE) model of instruction, where discussion is initiated by the teacher, the student responds, and the teacher evaluates the student's response. This change of roles in Period 2 from initiator with information to a supportive respondent suggests a modification in a traditional IRE pattern of instruction (Quinn et. al, 1983). Davie and Wells (1991) have suggested that an important "consequence" of a shift to more supportive behaviors by the teacher is increased teacher interaction with students. In this case, however, the instructor believed that it was increased student interaction with him, the instructor, that prompted him to change his role from the initiator of electronic messages to the respondent to electronic messages. This observation lends further credibility to the hypothesis that the increased growth in student messages was due to a mandatory reflection assignment and was not caused by a shift in the instructor's behavior. Rather, it is plausible that the consequence of increased student participation may have precipitated a change in the instructor's role from an initiator to a respondent.

4. Students who did not join the listserv

Up to this point, the focus of the case study has been on the instructor and 29 students who participated on the listserv. But what about the nine students who never went online? Two focus groups (a total of 8/19 students) were conducted during the semester. When asked about their reasons for not joining the listserv, several claimed that, "I am still pretty traditional and would go to the books first to look something up on insects," and "I haven't used the listserv yet because I could probably get the answer elsewhere." Another student added: "I only like to put valuable info on the listserv. I hate to waste people's time with something they might not be interested in so when I couldn't think of a really valuable reflection, I didn't post anything."

The seven remaining students that did not overcome their "activation barrier" may have hesitated from posting messages for a variety of reasons, such as poor typing skills, reading comprehension, lack of time, or the degree of convenience of system access (Althaus, 1997). A lack of confidence or greater need for comfort with technology may have been another reason for not posting, as implied by one member of the class from the questionnaire: "I enjoyed getting the e-mail and reading what everyone said but was inhibited about writing because SO MANY people would be getting it, seeing my stupid question, etc." The instructor recognized this issue:

  ... it's analogous to the way a class progresses and there may be a
  person by the middle of the semester who finally raises their hand
  and responds in class. So for some people it's simply a fact of
  that they've just gotten their e-mail and for other people it just
  takes a certain amount of risk taking; feeling comfortable; wanting
  to expose themselves in that public forum.

In a particularly unusual case, one student from the focus group was an active online user in another course but chose to refrain from participation and just read other students' messages on the listserv. She was a part of the focus group, and was asked to elaborate on why was there a difference in her online participation? "[Because] we are submitting many of our assignments that way [on the other course's listserv] and getting information for the assignments ... [For example,] one assignment was for everybody to find an interesting web site that had something to do with teaching." This student's remark suggests that one reason for the difference in her online participation between the two courses may have been the use of mandatory assignments.

CONCLUSION

The long range questions in this article are: How can instructors motivate students to participate in computer-mediated discussion? If they do participate, how can we assess the quality of discourse?

There are reasons to believe that listservs can augment classroom discussions productively. According to several researchers (Harasim, 1987a; Althaus, 1997; Quinn et al., 1983), listservs have the potential to enhance communication because they are independent of the constraints of place and time that are present in classroom settings. In this course, communication appeared to be enhanced since the length of student messages was longer on the listserv than responses in class, and for some students, there appeared to be more time for reading, reflection, and composition on the listserv than during class. The listserv clearly supported additional discussion about insects outside of the classroom.

Examining the electronic discussions themselves, student listserv activity was markedly greater in the second half of the course, referred to as Period 2 throughout the study. By the end of Period 2, a greater number of student participants joined the listserv; more student messages were posted on average per week, and several students overcame their "activation barriers." In the first half of the course, the teacher was initiating most electronic discussions and obtaining sporadic student response. In Period 2, however, the evidence shows that the instructor's messages paralleled student's messages, and after reading the text of these messages, it became clear that his messages were in response to students' questions. Figure 4 depicts this incongruous pattern in Period 1 compared with a roughly parallel pattern of teacher and student interaction in Period 2. All of these observations suggest that the listserv was becoming more student centered. The most likely indicator for this shift from a teacher centered to a student centered listserv was a mandatory assignment. This assignment was an online exercise that produced a burst in student activity, followed by only a gradual tapering off of activity for the rest of the course as represented in Figure 2. This pattern was present even with student responses to the assignment subtracted out of the total number of student postings per week as represented in Figure 3, indicating that the initial burst triggered a sustained level of interest in posting.

Associated with the sustained growth in student activity was an increase in multi-participant student discussions and an increase in student messages that contained reference citations. Multi-participant discussions and reference citations certainly do not capture all of what we intuit by "rich communication," but they do provide interesting variables to assess the quality of online discourse and perhaps present implications for expanding the degree of "information richness" for electronic media. Thus, the quality of student interaction and discourse on the listserv appeared to be more rich and more student centered by the end of the course.

RECOMMENDATIONS FOR TEACHERS

The benefits of computer mediated communication for inservice and preservice teachers have been well documented (Waugh, 1996; Thompson & Hamilton, 1991; Sunal & Sunal, 1992; Russet, 1995; Casey, 1998; Zimmerman & Greene, 1998); however, most of these studies have investigated the effect of the process of electronic communication on some product: typically student achievement or the instructor's pre-determined goals for the course. It is rare that a study examines the effect of an intervention on the process of communication. This study explored the effect of a mandatory assignment on the evolution of an online community of preservice and inservice teachers. In a previous study, 90% of online students felt that participation on such listservs should be entirely optional (Althaus, 1997), and there have been reports that unstructured listservs can be successful (Piburn & Middleton, 1998), but Althaus cautions that when left to choose for themselves, few students would participate freely on listservs. Two workable compromises to the mandatory/optional assignments question for teachers may be (a) to allow students a choice among alternate assignments that draw upon the same critical thinking skills or (b) to require each student to use the system at least once at the start of the course. Quinn et al. (1983) suggested that the first online assignments should include getting an e-mail account, logging on to the system, and sending a simple message to the instructor. Even though Session 2 of the course package had such an assignment, it was not counted towards the final grade and correspondingly, there was no significant burst in online activity. To encourage greater online activity, however, such assignments may need to be mandatory and contribute towards a final grade.

This study illustrates that even a single well chosen mandatory assignment can be effective in boosting and sustaining and increase in student participation, to the point where it may even motivate students to overcome their activation barrier. One student who did overcome the activation barrier because of the mandatory assignment concluded that "Once it was used by a greater percentage of the class it became a very effective tool for verifying assignments, information, and sharing of resources." This study supports the strategy of including one or more mandatory assignments to act as a catalyst for growth in student-centered computer mediated discourse.

The significance of this study for inservice and preservice teacher educators is that the findings revealed, contrary to some literature, that the mandatory assignment contributed to the development of a student centered electronic community. Based on the evidence, it appears that the mandatory assignment boosted the number of student participants and the number of student messages. Access, comfort with computer mediated communication, and shift in the teacher's behavior did not appear to be the catalyst for the growth of the listserv. While the immediate effect of a mandatory assignment may not be a surprising outcome, it was not necessarily expected that the increase in student participation would be sustained beyond the mandatory assignment. The implication of this finding for teacher educators is that a single mandatory assignment may have positive effects for the development of an online community. Rather than encourage only teacher initiated postings that may short circuit student discussion (Mazzolini & Maddison, 2003), this study also recommends mandatory assignments as a potentially effective strategy for those teacher educators interested in boosting and sustaining student participation on a listserv. Parameters that were not investigated in this study but may have been associated with the apparent effect of the mandatory assignment were the nature of the task itself (to share a prior experience), the degree of familiarity of the participants with each other (the assignment was midway through the term), and the nature of the instructor's responses to the assignment once it was posted (responding in support of a particular student as opposed to a general broadcast or an evaluation of the student message). Thus, teacher educators who are interested in the development of an online community may wish to consider at least one mandatory assignment.

FURTHER STUDY

Creating the social organization of a listserv takes careful planning. Too many networking communities have fallen silent, as electronic ghost towns (Riel & Levin, 1990). In a comparison of several telecourses, Riel and Levin found that the nature of the task was an important variable in determining the success of the electronic community. They reported that successful electronic communities provided a goal and a timeline for the end-product compared with unsuccessful electronic communities that were general and unspecified about goals and timelines. The fact that the mandatory reflection assignment had a clear objective and an expectation for completion by the end of the course may have also contributed to the success of the listserv in this course.

Other variables that were not investigated in this study but may have contributed to the development of a student-centered listserv were the nature of the task itself, the degree of familiarity of the participants with each other, and the nature of the instructors' responses to the assignment once it was posted. Thus, questions for further investigation include:

  If students were asked to write a group or individual report on the
  biology of an insect rather than reflective teaching (Walther,
  1992), would it have elicited the same degree of change in the
  quality and quantity of student and instructor computer-mediated
  discourse?

  Would the same pattern of student responses to the assignment appear
  if the mandatory assignment was given earlier in the semester when
  students and instructors were strangers (Weir, 1992; Caggiano,
  Audet, & Abegg, 1995) instead of half-way through the course, or if
  the assignment was given in a course that was not coordinated with a
  classroom component?

  How would the sustained growth in student responses to the listserv
  be affected if the instructor's responses to messages tended to be
  evaluative and critical, rather than supportive in tone, delayed
  rather than immediate?

These questions may provide suggestions for further research on the best ways to foster participation and richness in computer-mediated discourse.

Table 1 Modes of Interaction on the Listserv

Direction of           Listserv Messages
Communication          Period 1  Period 2

I[right arrow]S        97%       77%
S[right arrow]I
I[right arrow]C
S[right arrow]S
"Multi-participant"     3%       23%

Note: I = instructor, S= student, C= class. 97% of all listserv messages
in Period 1 could be characterized as unidirectional; that is,
instructor to a student, a student to the instructor, instructor to the
class, a student to another student.

Note

The author would like to acknowledge the extended time and effort contributed by Dr. John Stoffolano and Faith Thayer in data collection and reflection on the events of their course.

References

Althaus, S. (1997). Computer-mediated communication in the university classroom: An experiment with online discussions. Communication Education, 46, 159-173.

Caggiano, M.E., Audet, R.H., & Abegg, G. (1995). A pilot study of an electronic community of interdisciplinary secondary science teachers. Paper presented at the Annual Meeting of the National Association for Research on Science Teaching, San Francisco, CA. (ERIC Document Reproduction Service No. ED382473)

Casey, J. (1998). TeacherNet: Online k-12 classrooms now ready for online student teachers. Technology and Teacher Education Annual, Proceedings of the International Conference of the Society for Information Technology and Teacher Education (SITE). Washington, DC.

Daft, R.L., & Lengel, R.H. (1984). Information richness: A new approach to managerial behavior and organization design. Research in Organizational Behavior, 6, 191-233.

Davie, L.E., & Wells, R. (1991). Empowering the learner through computer-mediated communication. The American Journal of Distance Education, 5(1), 15-23.

Foulger, D.A. (1990). Medium as process: The structure, use, and practice of computer conferencing on IBMs, IBMPC computer conferencing facility. Unpublished doctoral dissertation, Temple University, Philadelphia.

Harasim, L. (1987a). Computer-mediated cooperation in education: Group learning networks. Proceedings of the Second Guelph Symposium on Computer Conferencing, (pp. 171-186), June 1-4. Guelph, Ontario, Canada: University of Guelph.

Harasim, L. (1987b). Teaching and learning online: Issues in computer-mediated graduate courses. Canadian Journal for Educational Communication, 16(2), 117-135.

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

Harasim, L., & Yung, B. (1993). Teaching and learning on the Internet. Burnaby, BC: Department of Communication, Simon Fraser University.

Harasim, L., Hiltz, S., Teles, L., & Turoff, M. (1995). Learning networks. Cambridge, MA: MIT.

Hiltz, S.R., & Turoff, M. (1994). Virtual classroom plus video: Technology for educational excellence. Educational Multimedia and Hypermedia, 1994, Proceedings of ED-MEDIA 94--World Conference on Educational Multimedia and Hypermedia, (pp. 26-31), Vancouver, BC, Canada.

Honey, M., & Henriquez, A. (1993). Telecommunications and k-12 educators: Findings from a national survey. New York: Center for Technology in Education, Bank Street College of Education.

Hsi, S., & Hoadley, C.M. (1997). Productive discussion in science: Gender equity through electronic discourse. Journal of Science Education and Technology, 6(1), 23-36.

Karayan, S.S., & Crowe, J.A. (1997). Student perceptions of electronic discussion groups. T.H.E. Journal, 24(9), 69-71.

Mazzolini, M., & Maddison, S. (2003). Sage, guide, or ghost? The effect of instructor intervention on student participation in online discussion forums. Computers and Education, 40, 237-253.

McComb, M. (1994). Benefits of computer-mediated communication in college courses. Communication Education, 43(2), 159-170.

Piburn, M.D., & Middleton, J.A. (1998). Patterns of faculty and student conversation in listserv and traditional journals in a program for preservice mathematics and science teachers. Journal of Research on Computing in Education, 31(1), 62-77.

Quinn, C.N., Mehan, H., Levin, J.A., & Black, S.D. (1983). Real education in non-real time: The use of electronic message systems for instruction. Instructional Science, 11, 313-327.

Riel, M.M., & Levin, J.A. (1990). Building electronic communities: Success and failure in computer networking. Instructional Science, 19, 145-169.

Romiszowski, A.J., & de Haas, J.A. (1989). Computer mediated communication for instruction: Using e-mail as a seminar. Educational Technology, 29, 7-14.

Russet, J. (1995). Using telecommunications with preservice teachers. Journal of Computers in Mathematics and Science Teaching, 15(1/2), 65-75.

Steinfield, C.W., & Fulk, J. (1986). Task demands and managers' use of communication media: An information processing view. Paper presented at the Meeting of the Academy of Management Organization Communication Division, Chicago, IL.

Sunal, D.W., & Sunal, C.S. (1992). The impact of network communication technology on science teacher education. Journal of Computers in Mathematics and Science Teaching, 11, 143-153.

Thompson, A.D., & Hamilton, J. (1991). Patterns of use of an electronic communication network for student teachers and first year teachers. (ERIC Document Reproduction Service No. ED334180)

Trevino, L.K., Lengel, R.H., & Daft, R.L. (1987). Media symbolism, media richness, and media choice in organizations. Communication Research, 14(5), 553-574.

Walther, J.B. (1992). Interpersonal effects in computer-mediated interaction. Communication Research, 19(1), 52-90.

Waugh, M. (1996). Group interaction and student questioning patterns in an instructional telecommunications course for teachers. Journal of Computers in Mathematics and Science Teaching, 15(4), 353-382.

Weir, S. (1992). Electronic communities of learners; Fact or fiction. (ERIC Document Reproduction Service No. ED348990)

Zimmerman, S.O., & Greene, M.W. (1998). A five-year chronicle: Using technology in a teacher education program. Technology and Teacher Education Annual, Proceedings of the International Conference of the Society for Information Technology and Teacher Education (SITE). Washington, DC.

SAMIA KHAN

University of British Columbia

Vancouver, British Columbia, Canada

[e-mail]