Participation Patterns in an Electronic Network-Based Tele-Task Force Instructional Project:
The Zero-g World Design Project

Michael Waugh
Pia Bombardier

311 Education Building
1310 South Sixth Street
Champaign, Illinois 61821

Telephone: (217) 244-7005

Abstract

This paper will describe an analysis of the nature of individual participation during the Zero-Gravity World Design Project (Zero-g), an electronic network-based project activity conducted during the 1991-92 school year. Participants from two different electronic networks, the Internet and FrEdMail, took part in the project. Various questions were asked about the organizational nature of the project and the data collected were examined using quantitative methodologies. How was the project organized? Who participated? Who didn't? What was the level of active participation? Why? What relationship existed between home network affiliation (i.e., Internet or FrEdMail) and level of involvement and response rate? Answers to these and other questions offer insights concerning effective network-based project management, and have important implications for structuring educational uses of electronic networks.

Introduction

The Zero-g project took place during the 1991-92 school year. It can be described as a Tele-Task Force project, a term employed by Levin, Waugh, Chung & Miyake (1992) to describe network mediated activities in which the participant group comes together based upon their common interest in the project topic but members of the group are under no obligation to participate at any fixed level on the project or any subsequent post-project activities. Thus, the participants are simply volunteers who were attracted when the project coordinator announced the existence of the project and solicited involvement and input from interested individuals. The Zero-g project was organized and conducted in collaboration with all interested parties and the specific sequence of events and outcomes were not pre-planned but rather were developed in collaboration with other project participants as the project progressed (see Waugh, Levin & Smith, 1994a, for a further description of this type of project organization). The project participants came from a variety of different geographical locations and local network environments. The information regarding the specific local network environments was not preserved, however two distinct sub-groups were readily identifiable. One sub-group used the FrEdMail network which possesses in a terminal/host architecture. The other sub-group used the Internet which interconnects a variety of mainframe-based terminal/host or client/server architectures.

Project Initiation

The following message was sent to several listservs and netnews groups on the Internet and the bulletin boards on FrEdMail to announce the Zero-g project activity and to solicit interested participants. The original message was sent to approximately ten such public forums, however it became evident during the project that the announcement was forwarded by third parties to a considerably larger number of other "public" and "private" destinations. The total number of individuals who may have seen the project announcement cannot be known with certainty.

The original message is included below exactly as it was sent with the exception of the addressee list which has been edited to simplify the header for clarity.

Message 1

The purpose of this project announcement was to paint a broad picture of the nature of this project, to solicit involvement from the recipients and to inform the participants that their input was desired in helping to shape the activity.

Project Organization

The Zero-g project began when the project coordinator sent Message 1 and solicited involvement from others on the network. The project evolved as the project coordinator interacted with these participants throughout the duration of the project. The series of activities which define the project were conducted over the course of a year. These five activities are the Collision Challenge, the Design Challenge, the Christmas Mini-Challenge, the School Design Challenge and the Five Same and Different Challenge. These five activities are briefly described below.

Collision Course Challenge: Participants were to think about what they would do if they were floating down a corridor in zero gravity and another person was floating toward them from the opposite direction on a collision course. Students were challenged to develop a plan for how would they avoid a collision.

Design Challenge: The participants were to design a recreational activity (either take an existing, known activity like soccer and modify it for space or create a new type of activity) or a food design (eating in space, cooking in space, etc.).

Christmas Mini-Challenge: Participants were to think about how Christmas would be celebrated in zero-g. What might be similar to what we do here on earth and what might be different?

Zero-g School Challenge: Participants were to design a school that would be appropriate for a zero gravity environment. How would a school be different in zero-g?

Five same & Five different Challenge: Participants were to come up with five things that would be the same and five things that would be different about living in zero gravity.

Project participants

Participants were invited to join the project activities regardless of their network affiliation and individuals from a variety of networks participated. The K-12 participants in the Zero-g project were predominantly users of the FrEdMail network. Some of the K-12 participants had direct access to the Internet but the majority were FrEdMail users. The higher education participants were predominantly users of the Internet network. For the Internet users, the specific characteristics of their local networks is unknown. However, these participants all utilized networks which were interconnected directly to the Internet for email exchange.

Project participants were either individuals (teachers, professors, K-12 students or university students), or K-12 class groups. Class groups were often represented by single individuals (teachers or others) mediating their involvement in the project.

The data that describe the entire group of project participants includes the participants from a wide variety of networks, both commercial and academic (.i.e., Compuserve, BITNET, etc.). However, a smaller sample of 18 participants is analyzed in greater detail and this dataset is composed of only individuals from the FrEdmail and Internet networks.

FrEdMail Network

FrEdMail is a distributed electronic network populated primarily by students and teachers in the K-12 schools. Most of the computers which serve as hosts or nodes in the network are Apple IIe and IIgs computers. These computers serve as "hubs" for local users and provide email and bulletin board services. A local user may use any microcomputer and a terminal emulation software package to dial in to the local FrEdMail node using a regular phone line. The FrEdMail nodes typically exchange email and bulletin board messages daily through telephone connections. Recently, FrEdMail has developed several gateway connections to the Internet to facilitate long distance message exchanges.

The Internet Network

The Internet is the largest electronic network in the world. It is a worldwide network of networks that interconnects smaller networks through the use of common communication protocols. The Internet is composed of millions of host or server computers that are directly connected to one another by dedicated telecommunications circuits. In addition, many millions of users also establish periodic connections to the Internet via temporary telephone connections. Thus, the nature and capability of a user's local network "environment", i.e. their specific hardware and software capabilities, varies widely. The two major types of system architectures are terminal/host and client/server systems. In terminal/host systems, most of the software resides on the host computer and the user would connect to the host to run the host-resident software. In client/server systems, part of the software exists on the server and part exists on the client computer. These complementary software packages are designed to take maximum advantage of the capabilities of both computers. Communication between the client and host computers is accomplished only when necessary. For users who connect to the Internet via telephone lines, client/server systems offer a more efficient means of connection and information exchange.

As the different project activities were conducted, the participants joined or left the activities according to their preference. The participants were invited to discuss the structure of the project activity and assist in shaping the activity throughout the term of the project. The size of the group of project participants varied throughout the term of the project. Participants regularly requested that they be added or dropped from the group list. The largest number of participants to be involved at any point in the project was 106. The participants' levels of involvement varied widely. Some participants were "active", meaning that they contributed at least one activity-related message to the group. Other participants were "passive", meaning that they continued to receive all project announcements and correspondence but did not make contributions to the group.

Figure 1 illustrates the "flow" of the project across the academic year as measured by messages sent by participants to the whole group. The challenge labels indicate the sequence and approximate timing of the occurrance of the challenges during the adademic year. Participation in the project began slowly in the fall period and increased during the spring period which is a common pattern in activities of this type (Levin, Waugh & Kim, 1990).

Figure 1. Zero-g World Design Project activity

Methodology and Results

The remainder of this paper describes analyses of several variables which relate to the level of project involvement by participants. Summary statistics for the entire project will be presented first and these will be followed by a more in-depth analysis of a smaller group of active project participants. These analyses are quantitative. However, whenever possible, the findings are elaborated by information supplied through participant interviews.

These results are presented in order to describe what occurred during the Zero-g project, and in hopes that they might offer some information for those attempting to conduct network-mediated projects with similar characteristics (i.e. year-long in duration, Tele-Task Force type and with participants from multiple network environments). However, because of the voluntary self-selection of project participants and relatively small sample size it is impossible to determine how representative this group might be and generalizability of results from these analyses should be made with caution.

Over the term of the Zero-g project, a total of 106 individuals responded to the project announcement, indicated an interest in the project activities, and asked to be placed on the project mailing list. The project was conducted by participants addressing messages to the group email "reflector", an electronic mail address for the group which "echoed" or forwarded the correspondence to each name contained on the participant list. Thus, any messages addressed to the email reflector address was received by all participants which resulted in an electronic mail-based conferencing mechanism.

The group of project participants was dynamic with some individuals joining throughout the term of the project and others leaving at various times. Thus, membership in the project was not stable throughout the year. During the project, this total group of 106 participants generated 223 messages related to the project (not including the messages they sent to join or leave the project). Sixty-five of these messages were sent during the first semester (August-December) and 158 messages were sent during the second semester (January-June).

In reviewing the first semester activities prior to the beginning of the second semester, several questions emerged concerning the nature of the Zero-g participant group. Which of the participants currently on the project list, still wished to continue with the project during the second semester? Which network did they use? How active had they been so far? Was there any relationship between a participant's activity level, network residence and the speed with which they might be able to contribute to the project activities?

Active participation during the first semester

During the first semester (Fall, 1991), 65 participants registered for the project by requesting that their email addresses be included in the project email reflector list. Of this number, only 17 actually became active (contributing one or more messages). Thus, 74% of the registered participant group (48/65) were not active during the first semester. Because of this, an attempt was made to determine how many project participants wished to remain on the project participant email reflector list during the second semester. In order to answer this question, the project coordinator sent out a "ping" message (Waugh, Levin & Smith, 1994b), a short message which asked all members of the project group to reply and indicate whether or not they wished to remain on the project mailing list (see Message 2). This ping message can be considered to be representative of any particular coordination message sent from the project coordinator to the participant group requesting a response from them, and so the responses to the ping message were analyzed to determine the characteristics of the respondent group and their associations, if any.

Message 2

Active participation during the second semester

At the beginning of the second semester (Spring, 1992) when the ping message was transmitted, there were 93 project participants on the project email reflector list. Thirty-five project participants (38%) responded to the ping message. Fourteen of these individuals (40%) had been active project participants during the first semester. Of the 35 individuals who responded, 25 (71%) asked to remain on the reflector list and 10 (29%) asked to be removed from the mailing list. Of the 25 who wished to be kept on the mailing list for the second semester, 16 (64%) became inactive. Of the nine who became active during the second semester, five were on the Internet and four were on FrEdMail. In addition to the nine individuals who responded to the ping message and became active during the second semester, 18 additional participants became active during the second semester. This unexpected participation was possible because no email address was ever removed from the project email reflector unless the project coordinator received specific instructions from an individual participant requesting that such action be taken. Even though individuals were not actively participating, they continued to receive all messages sent to the project group unless they specifically requested to be removed from the project's electronic mail reflector list.

Selecting an "active" sample for further analysis

The 35 participants who responded to the ping message were further analyzed to determine which of them were active during any challenge activity in the project and this yielded a group of 18 individuals. Thus, the sample group was composed of 18 active participants, all of whom had responded to the ping message.

Factors related to participant activity level

The remaining questions addressed in this study can be summarized as follows: Does network affiliation (the network used by an individual, i.e. their network environment) have any relationship to either speed or magnitude of participants' involvement? Speed of response was operationalized in this study as the order in which the participants responded to the ping message and this set of responses can be considered to be a random sample from the distribution of all such possible response sets. Although the individual's response times would vary across successive interactions of this type (ping and responses), this sample should be indicative of the general response capability within this active participant sample. Magnitude of response was operationalized as the number of messages contributed by each participant throughout the term of the project excluding any administrative requests to join or leave a project activity.

If it is assumed that all participants are equally motivated regarding the project, then variations in their response times (examined as their relative response order to the ping message) and in the magnitude of their responses must be influenced by other factors. One likely factor is network environment, i.e. the hardware and software tools which facilitate their communication process. This study provided an opportunity to compare those participants working in the FrEdMail environment with those using the Internet.

Specifically, the following questions were addressed:

1. Is there a relationship between the time required for an individual to respond (ping response order) to a message generated by the project coordinator and their network affiliation?

2. Is there a relationship between the time required for an individual to respond (ping response order) to a message generated by the project coordinator and the total number of messages they contributed to the project?

3. Is there a relationship between the number of messages sent by participants and their network affiliation?

Question One

To address question one, a Spearman Rank Order correlation was calculated on the response order and network affiliation data shown in Table 1. Response order was selected as the variable to examine (rather than response time as indicated by the time stamp on individual messages) because of the difficulty involved in establishing exact time across user machines and networks and in determining when the time stamp is affixed to a message. Additionally, the purpose of this analysis is to examine relative response order as an index of involvement and the practical issue is when the message is received rather than when it was generated.

Although the magnitude of the correlation between network affiliation and response order is not particularly high (r=.47), the pattern revealed by Table 1 shows that a relationship does exist between a participant's network and the order in which they responded to the ping message. Information obtained through follow up interviews with the three individuals who were exceptions to the general pattern illustrated by the other participants shown in Figure 1 (#1, #12, #17) indicated that their placement in the order of responses was influenced by chance and confusion. Participant #1 typically checked FrEdmail infrequently, perhaps once a week or less often, and by chance merely happened to call immediately after the ping message had been sent. Participants #12 and #17 were individuals at the University of Illinois affiliated more closely with the project who felt that the ping message was not intended for them because their continued participation should have been obvious to the project coordinator. However, both individuals independently decided at a later time that they should send a response and thus their responses were not sent as quickly as was possible.

Table 1


Active/Ping respondent's response order broken down by network affiliation


Response Order Internet FrEdMail

1 *

2 *

3 *

4 *

5 *

6 *

7 *

8 *

9 *

10 *

11 *

12 *

13 *

14 *

15 *

16 *

17 *

18 *

Note 1: '*' indicate net affiliation of respondent in ordered list.

Question Two

To address question two, a Spearman Rank Order correlation was calculated on the participant's response order and number of messages contributed. This analysis produced a correlation of -.21 and this value along with an examination of a plot of the data, indicate that little relationship exists between the order in which the individuals replied and the magnitude of their project contributions.

While it might be tempting to interpret this as "a project participant's ability to quickly interact is not related to their level of project involvement" (and thus, for example, the nature of an individual participant's personal schedule and hardware/software environment are not important), this specific finding does not support such a conclusion. This finding illustrates that relative response rate is not related to the magnitude of an individual's contributions. Based on this finding however, it is impossible to determine how ease of interaction or speed of interaction might influence extended participation in project activities of this sort.

In this particular project, the magnitude of a participant's contributions bore little relationship to the order in which they responded. This finding may mean that participants from a wide variety of networks (with varying technical capabilities) have the possibility for meaningful collaboration, i.e. they could work together (with approximately equal levels of contribution) on joint activities. However, we have no notion as yet of their relative probability of success in such joint activities. Many factors are likely to play critically important roles in the success of joint activities across networks which vary according to the technical characteristics and interface environments they offer their users.

Question Three

To address question three, another Spearman Rank Order correlation was calculated on the data shown in Table 2. Again, the correlation coefficient was not particularly high (r=.38) however, a pattern is evident in the data which is worthy of discussion. The data in this table verify a finding reported by Levin et al. (1992) relating to institutional patterns of network involvement. Briefly stated, different institutions possess different characteristics such that each possesses a unique communications profile, i.e. there are certain times of day/week/month/year when members of the institution are more likely to have time to communicate with others; and institutional affiliation may also influence the content of the communications. In this case, the data illustrate that the type of contributions made by those participants on the Internet were different than those made by the FrEdmail participants (see Figure 2). Relatively more Internet participants contributed fewer messages each (nine participants contributed five or less messages with only two contributing more than nine messages), whereas relatively fewer FrEdmail participants each contributed larger numbers of messages (five participants contributed six or more messages with only two contributing two or less messages). One likely interpretation of this pattern is that the Internet participants were largely individuals, whereas the FrEdmail participants were more often K-12 teachers who were coordinating their classroom contributions through their personal FrEdMail accounts. Thus, this phenomenon may have little to do with the actual network affiliation of an individual and might be more influenced by the nature of the participant's roles and institutional constraints, i.e. autonomous adults versus adult mediators organizing the contributions of classrooms or other small groups.

Table 2

Active/Ping respondent's contribution levels broken down by network affiliation

Messages Contributed Internet FrEdMail

1 3 1

2 2 1

3 1 0

4 2 0

5 1 0

6 0 1

7 0 1

9 1 0

10 0 2

13 1 0

28 0 1

Note 1: Cell contents are the number of participants OUT OF 18 active participants who responded to the ping message.

Figure 2 illustrates the data from Table 2 by grouping participants as contributors of either 5 or less messages, or 6 or more messages.

Figure 2. Message contributions broken down by network affiliation

Discussion

Response time and network affiliation

Based on the data shown in Table 1, a clear relationship exists between the order of participant responses and their network affiliation. This is probably due to several factors. One such factor is the fact that FrEdMail nodes do not exchange messages as frequently as is the case in the Internet. FrEdMail uses a store-and- forward procedure based on daily (typically) telephone connections. Internet host and server machines are directly connected to rest of the network and electronic mail is typically sent immediately or within a matter of minutes. Another factor is that the teachers and other FrEdMail-based participants did not check their email quite as often as did the Internet-based participants in this project. This latter phenomenon is probably due, in part, to the fact that the FrEdMail system software is not as easy to use as many of the newer client/server packages used by the Internet participants (Eudora, NUPOP). However, it is also partly due to the nature of the institutional constraints associated with the K-12 environment. Simply put, teachers and students are not typically free to interact with an electronic network on as frequent a basis as the participants in this project who used the Internet. While this "opportunity to frequently interact with the network" will vary widely within the K-12 environment and within the university environment, as a general rule the K-12 schools are generally more structured than the university environment and would tend to offer less blocks of time for students and staff to access computers and networks. In addition, most of the project activities were conducted "off-line" in the classrooms and longer periods were required to conduct these portions of the activity. A consequence of this is that the "on-line" portions of project activities of this type will tend to be farther apart and communications less frequent.

The relevance of this finding is not that participants on one network are faster than those on another but rather that networks and the individuals which compose them will vary in their rates of interaction (likely due to many factors) and project coordinators who wish to work with internetworked participant groups should be aware of this and allow sufficient flexibility in a project timeline to accommodate such differences.

Response time and magnitude of contributions

No relationship was found between the order of response to the ping message and the overall number of messages a participant contributed. Thus, the participant's relative speed of response in this project was not a crucial factor associated with their level of participation. However, this does not mean that speed of response is not an important factor in broadening and sustaining participation-- a determination of this sort is beyond the scope of the data in this study.

In the Zero-g project, the FrEdMail participants demonstrated a willingness to participate in a project of year long duration and to follow through with all aspects of the activity. However, this behavior does not allow us to predict their future involvement and the more difficult it is for teachers/students to communicate and collaborate in instructional activities of this type, the less likely will be their continued participation. Thus, factors which influence ease of communication (perhaps measured as the individual's ability to quickly respond) such as network connection type and hardware and software--local network environment variables--should be explored to determine their long term impact on magnitude of contributions and continued participation.

Magnitude of contributions and network affiliation

The type of involvement by participants varied by network affiliation (see Figure 2). Fewer FrEdmail participants contributed more messages each and more Internet participants each contributed fewer messages to the project activities. Typically, the Internet participants were individuals contributing information to the project on their own and at their own pace, whereas the FrEdmail participants were teachers contributing information on behalf of their classes or some smaller project group. This latter type of involvement by the FrEdMail participants tends to slow down the communication and network interaction and it also tends to produce contributions which are complex in that many individuals have contributed to a single message. As software and hardware and skill levels continue to evolve and the disparity in technical capability across networks diminishes, then perhaps these differences in messaging patterns by participants across networks will disappear. However, characteristics associated with the institution of K-12 schools will also influence the nature of interactions between network resources and classrooms and these characteristics may play an even more significant role in influencing the rates of interaction and the types collaborations which are possible through electronic networking.

Additional findings

One surprising finding from this project was the degree of inactive participation or passive observation which occurred. These participants continued to "observe" the project activities yet made no active contributions to it. The project had a total of 106 individuals join throughout its duration. Of these 106 individuals, 44 (17 first semester, 27 second semester; 42%) became active participants during the project. Thus, in a global sense, the project lost 58% of its potential participant group. Focusing only on the second semester, the numbers are fairly similar. Of the 93 registered participants at the beginning of the second semester, 35 (38%) responded to the ping message. Sixty-two percent did not respond. Of the 35 respondents, 25 (71%) wished to continue and 10 (29%) wished to quit from the project. Although these numbers vary a bit, they hover around a 1/3 : 2/3 ratio. One interpretation of this is that project coordinators should expect that only about one third of those individuals indicating a willingness to join or participate in their projects will become active participants. Another possibility is that these numbers are unique to this project and not at all generalizable. In the absence of additional data, Tele-Task Force project coordinators should, at the very least, expect some and perhaps significant inactive participation/passive observation throughout the duration of their projects.

Several reasons might account for the inactive participation seen during this project. This is a time when many educators are interested in how this medium can be used for instruction. It is possible that many such individuals join as many projects of this sort as possible in order to observe and learn from these efforts. Thus, their intention might never have been to become active but rather simply to observe the activity. Another reason might have been that since the project was a Tele-Task Force project and did not begin with well defined and explicit goals, as the project unfolded the individuals realized that they could not adapt the topic to their settings or that they, themselves, had little expertise to contribute to the effort.

Another surprising finding from this study was the poor response to the ping message sent at the middle of the project. Sixty-two percent of the registered project participants failed to respond to the ping message. However, in addition to the nine individuals who responded to the ping message and continued to be active during the second semester, eighteen others who did not respond to the ping message became active during the second semester-- two thirds of those who actively participated during the second semester had not responded to the ping message! This is surprising because the request for input came from the project coordinator and required very little action on the part of the participants. Did the participants not receive the ping message or was their lack of compliance willful or due to a set of factors which will persist in future activities of this type?

It would be difficult to determine all of the reasons why participants failed to respond to the ping message, but an interpretation of the fact of their non-response is that project coordinators should not assume that the full membership of a network project will receive or correctly interpret any given message sent. How might this situation be addressed in hopes of improving the probability that administrative messages (or any message) will be read and acted upon? One strategy to address this problem would be the use of a header that clearly indicates that the communiqué is administrative/important for coordination. Also, the project coordinator could resend important messages to the specific group of non-responders until all respond. Although neither of these is desirable from the standpoint of making project management easier, neither is it desirable to lose significant numbers of project participants because they either did not receive a message, did not read it, did not read it closely enough, did not reply properly, experienced technical difficulties, etc. More work should be done to examine the relationship between project management strategies and participant engagement/retention in Tele-Task Force projects.

Implications for project management

The findings from this study offer several valuable suggestions for project coordinators wishing to conduct internetworked Tele-Task Force projects. These findings suggest that internetworked activities of this sort can occur and that participant groups from different networks can contribute equally toward the goals of the project. However, project coordinators should expect variability in the rates of project contributions. Project coordinators should also understand that the types of participation patterns will vary among users of different networks and this may be due to the nature of various specific institutional constraints on the participants themselves. Project coordinators should also expect low levels of active participation yet understand the importance of passive participation and the tendency for participants to migrate between the active and passive categories over the course of an extended network project. Lastly, project coordinators should never assume that any given coordination message will be acted upon. Critical coordination messages may need to be "broadcast" more than once over a period of time and ample time must be provided for effective coordination to occur.

References

Levin, J. A., Waugh, M. L.& Kim, H. (1990, April). Learning in Electronic Networks: Global and Local Activity Cycles. A paper presented at the American Educational Research Association Annual Meeting, Boston.

Levin, J. A., Waugh, M. L., Chung, H. K. & Miyake, N. (1992). Activity cycles in educational electronic networks. Interactive Learning Environments, 2(1), 3-13.

Levin, J. A., Waugh, M. L. & Stapleton, C. E. (1992). Coordination and mediation in educational networks: factors contributing to success and failure. Paper presented at the American Education Research Association Annual Meeting, San Francisco.

Waugh, M. L., Levin, J. A. & Smith, K. (1994a). Organizing electronic network-based instructional interactions: successful strategies and tactics, Part 1. The Computing Teacher, 21(5), 21-22.

Waugh, M. L., Levin, J. A. & Smith, K. (1994b). Network-based instructional interactions, Part 2: Interpersonal strategies. The Computing Teacher, 21(6), 48-50.

Acknowledgments

This material is based upon work supported by the National Science Foundation under Grant No. RED-9253423. The Government has certain rights in this material. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.