PATTERNS OF LEARNER-LEARNER INTERACTION IN DISTANCE LEARNING NETWORKS

Full paper for EDMEDIA2000 Annual Conference, Montreal, Canada, June 26-July 1, 2000 PATTERNS OF LEARNER-LEARNER INTERACTION IN DISTANCE LEARNING NETWORKS C. Candace Chou Interdisciplinary PhD Program in Communication and Information Sciences University of Hawaii Honolulu, Hawaii 96822 chou@hawaii.edu INTRODUCTION Abstract: This study examines the different patterns of online interaction between asynchronous communication and synchronous communication networks. Discussion transcripts were analyzed and coded using Bale s Interaction Process Analysis (IPA) model (revised and expanded version). The preliminary findings showed significant differences in the interaction between social-emotional and task-oriented contents. There was a higher percentage of social-emotional oriented interaction in synchronous communication. Whereas in asynchronous communication mode, a significantly higher percentage of online interaction were task-oriented. Furthermore, asynchronous discussion appeared to be mostly one-way communication and synchronous discussion showed evidence of two-way communication. Computer-mediated communication (CMC) systems have become an integral part of distance learning. Distance courses are conducted over CMC systems and learners interact with instructors or peers through computer conferencing. Studies on interaction over CMC networks have provided educators and researchers information on learners in terms of both intellectual growth and social development. Research findings on how interaction contributes to intellectual growth such as knowledge construction (Gunawardena, 1997), academic achievements (Hartman et al, 1995), and critical thinking (Newman, Webb, & Cochrane, 1995) indicates that cognitive development can be achieved over distance learning networks. Research on the social aspect of interaction also furthers the understanding of learner behaviors and dynamics in distance education (Johanson, 1996; McDonald & Gibson, 1998). The majority of research on learner interaction was conducted over asynchronous computer networks which are the primary media for distance education (Lewis et al., 1999). Few studies have been done on learner interactions in a synchronous communication network and even fewer research projects have been designed to compare both synchronous and asynchronous communication. Factors such as difficulty in coordinating meeting time, high cost in good quality synchronous communication technology, and tool stability may explain the under-utilization of synchronous CMC systems. Nevertheless, with the improvement in CMC technology and more affordable tools available, synchronous conferencing systems have become more common in distance learning environments. The problems of the previous research can be summarized as follow: a). lack of long-term study: most of the research uses previously unacquainted participants and short-lived group; b). lack of study on synchronous interaction: much emphasis has been placed on interaction over asynchronous communication networks; and c). lack of study on the comparison of asynchronous and synchronous interaction. Moore (1989) defines three types of interactions in distance education: learner-content interaction, learnerinstructor interaction, and learner-learner interaction. A comprehensive understanding of each type of interaction can help educators better plan learning activities and make use of the CMC technology for maximum effectiveness. Most of the research on online interaction draws conclusions from studies on asynchronous communication networks, e.g. electronic mail, listserv, web-based bulletin, BBS, etc. Little research has been conducted over synchronous networks. What are the nature and patterns of synchronous interactions? To what degree can learner interactions over synchronous learning networks contribute to learning? It is the purpose of this proposed study to examine learner-learner interactions in a collaborative online learning environment. The objectives for this study are: p. 1

1. To identify the patterns and developments of learner interactions over time, especially on task-oriented versus social-emotional oriented content; 2. to determine if there is a difference in the interactions when learners assume the roles of moderators and participants; and 3. to compare learner-learner interactions between synchronous and asynchronous communication networks. LITERATURE REVIEW Learners abilities to interact with the instructor, the peers, and the content can affect their performance in distance learning. Acker and McCain (1993) state that "interaction is central to the social expectations of education in the broadest sense and is in itself a primary goal of the larger educational process and that feedback between learner and teacher is necessary for education to develop and improve" (p. 11). Online interactions take into consideration the characteristics of the learners as well as the communication technology. The interactive features of the current CMC systems, such as two-way video and instant feedback, have provided more options for learner interactions. Gunawardena (1998) interprets interaction as the process through which negotiation of meaning and co-creation of knowledge occurs in a constructivist learning environment (p. 141). Wagner (1998) argues that interaction can serve as a means to an end of enhancing learning and performance. The design of the learning environments influences the results of learner-learner interactions. Scardamalia et al. (1992) found significant difference in the verbal scores from a Canadian standardized achievement test between elementary students who used Computer Supported Intentional Learning Environments (CSILE) to collaborate with teammates and those who used CSILE for individual study. CSILE is a multimedia learning environment that allows students to collaboratively contribute to one another s learning through co-construction of a knowledge database. As suggested by Martens et al. (1997), interaction is supported by the user interface that enables the learner to access and manipulate the objects in the knowledge base (p. 46). Good interface design provides students the opportunities to learn by interaction with their peers. One should not equate interaction with learning. It depends on whether the goal of an interaction task or activity is for the purpose of enhancing understanding of the subject matter or for improving interpersonal connections. Learner interactions require planning and structure in order to achieve the goal of active learning. Rohfeld and Hiemstra (1995) suggest tasks such as debates, guest lecturers/discussants, polling, brainstorming, or student-moderated discussions via CMC networks to increase student interactions for learning. The principles of student-centered discussion accord the students the responsibilities of facilitating online conversations. The process serves well for both cognitive and affective purposes. On the one hand, the process stimulates intellectual growth and enhances organizational skills; on the other hand, it improves the student connection at the interpersonal level (Paulsen, 1995). When the activities and tasks become an integral part of the learning process, learner interactions can be conducive to learning. Bales Interaction Process Analysis (IPA) (1950) has been utilized to examine small group interactions in both face-to-face and CMC groups (Rice & Love, 1987; Sorensen & McCroskey, 1977). IPA consists of twelve categories which fall into two main areas: social-emotional oriented and task-oriented contents. IPA was first developed in 1950 as the result of studies of group processes, e.g. who did what, who spoke to whom, and how interactions developed and changed (Bales, 1950). Bales IPA describes well the process of interaction and the dynamic in small groups. The twelve categories were modified for this study in order to better described the nature of online interactions. More details will be described in the following section. RESEARCH DESIGN RESEARCH QUESTIONS To further the understanding of online learner-learner interactions, the research questions will address learner interactions in the following areas: learning tasks, learning environments, learner characteristics, and the relationship with learning. The main questions are listed as follows: p. 2

A. Learner-learner interactions in social-emotional oriented contents versus task-oriented contents: How do online interactions develop over time in terms of social-emotional versus task-oriented content as outlined in Bales' Interaction Process Analysis? Is this pattern of interaction different between asynchronous communication mode and synchronous communication mode? B. Learning task: Will role-taking, such as assuming the role of a moderator or a participant, make a difference in online interactions? C. Learner characteristics: Do learner characteristics such as gender, attitude toward computer networks, experience in using computer, confidence in computer skills, affect the way they interact online? D. Learning environments: What are the similarities and differences in learner interactions via synchronous and asynchronous networks? METHODOLOGY This study used mainly qualitative methodology with the support of statistical analysis program. Content analysis was employed to examine learner-learner interactions and learning. Bales Interaction Process Analysis (IPA) with minor modifications was used as the basis of content analysis for learner interactions. The procedure was to systematically code student's weekly postings and online seminar transcripts in order to detect the interaction patterns. The research was carried out by surveying distance learners and analyzing student written data in a ten-week writing-intensive online course titled "Theory and Application of Computer-Mediated Communication." This course was delivered via a courseware called WebCT at the University of Hawaii in 1998. The data include weekly postings in the bulletin board (asynchronous data) and transcripts from a weekly seminar (synchronous data). In the weekly seminar, students met in a chat room in real time to discuss course-related topics from any locations that they could access a computer. Students were divided into five three-member small groups. Each group took turns to moderate one online seminar throughout the whole semester. Two researchers who specialize in instructional design worked on the coding of the data. The software program QSR NUD*IST was used for the coding of online transcripts. The unit of analysis was the sentence. A total of 5,015 text units (sentences) from the random sample transcripts were coded. The phi coefficient for intercoder reliability was.80. Bale s IPA was expanded from 12 categories into a total of 16 categories as shown in table 1. The original 12 categories can be divided into two main areas and four sub-groups. The two main areas are social-emotional contents and task-oriented contents. The social-emotional area consists of two groups: positive reactions and negative reactions. The task-oriented area consists of two groups that are attempted answers and questions. These areas represent the domain of interactions in small group discussions. The categories may not be comprehensive enough to describe the patterns of online interactions but the overall scheme does address most online interaction patterns. In his study on social-emotion interaction over computer-mediated communication networks, Rice (1987) expanded category 6 and 7 into two sub-categories. Category 6 on gives orientation, information, repeats, clarifies, confirms was divided into gives personal information (socioemotional) and gives professional information (task). Category 7 was divided into asks for professional information (task) and asks for personal information (socioemotional). For this study, category 6 and 7 were furthered divided into three sub-categories: gives/asks topic-related information (task), gives/asks personal information (socioemotinoal), gives/asks technical information (task). After the initial round of coding, the researcher found that the original category 6 and category 7 were too broad to reflect the actual online interaction patterns from the samples used for this study. Technical questions, topic-specific discussions, and personal information exchanges were frequently seen in the synchronous discussions and yet they were missing from the original IPA model. The researcher decided to include these new categories in order to better reflect the patterns of online interactions. DISCUSSIONS Coding online transcripts is a time-consuming process. In order to meet the submission deadline of the conference proceedings, only the data for the first research question is available for this report. The author will report the results of the data analysis for the rest of the research questions during the EDMEDIA2000 conference. The revised and expanded version of Bale s IPA is shown in table 1. There were significant differences in online interactions between asynchronous and synchronous communication modes using analysis of variances. In terms of each individual category, the differences were significant in the following categories: category 4 (gives suggestion, p <.01), category 5 (gives opinion, p <.01), category 7.2 (asks topic-related information, p <.05), category 7.3 p. 3

(asks personal information, p =.01), and category 11 (show tension, p <.05). In terms of the four sub-areas, i.e. positive reactions (p =.05), negative reactions (p <.05), attempted answers (p <.01), and questions (p <.01), the differences were significant. In terms of the two main areas of interaction, both social-emotional oriented contents (p <.05) and task-oriented content (p =.01) indicate the significant difference between the synchronous and synchronous communication networks (table 2). A closer look in category 4, 5, 7.2, 7.3, and 11 will provide a better understanding in the different interaction patterns between the two communication modes. First, there was virtually no interaction in category 4 (giving suggestion or direction) in asynchronous discussion mode. The mean sentence in category 4 is 0 for asynchronous discussion and 9.36 (SD: 7.49) for synchronous discussion. Category 9 (asking suggestions) is the counterpart of category 4. The mean sentence for category 9 is 0 too. It is understandable that in asynchronous communication most interactions focus on task-specific discussion and less on social-emotional oriented contents partially due to a lack of immediacy. Most students either internalized their attempts to ask for suggestions or email the instructor for advice. The time lag in asynchronous mode may have prevented students from asking or providing suggestions. Interestingly, the mean sentence in category 5 (giving opinions) is 247.33 (SD: 129.1) for asynchronous mode and 62 (SD: 42.05) in synchronous mode. It is a clear indication that in asynchronous mode, participants devoted most of the discussion to task-related discussion and less on social-emotional interaction. Although, there was no interaction in giving/ asking suggestion (category 4 & 9), the interaction in giving/asking opinion (category 5 & 8) was not scarce in asynchronous mode. Whereas in the synchronous communication mode, student felt it more imperative to ask and receive instruction or direction in order to participate fully. Synchronous communication mode also made it easier to provide immediate feedback to information seekers. The high standard deviation in synchronous interaction also reveals the unequal participation in online discussion. Some students were active in engaging in discussions and some students reserved most of their opinions until they were asked to say something. The researcher observed that when the students were divided into three-member small groups, the participation was more even among members of a small group. In addition, there were more social-emotional oriented discussions in synchronous mode that is indicated in the difference in category 7.3 and 11. In synchronous mode, participants asked more personal questions and revealed their frustration or needs for help without hesitation. Personal questions such as one s occupation, schooling history, professional training were given more air time in synchronous discussion. Table 1: Bale s IPA (revised and expanded) Code Category P value Social-emotional Area: Positive Reactions =.05* 1 Shows solidarity, raises other's status, gives help, reward =.013 2 Shows tension release, jokes, laughs, shows satisfaction =.015 3 Agrees, shows passive acceptance, understands, concurs, complies =.047 Task Area: Attempted Answers <.01 * 4 Gives suggestion, direction, implying autonomy for other <.01 * 5 Gives opinion, evaluation, repeats, analysis, express feeling, wish <.01 * 6 Gives orientation, information, repeats, clarifies, confirms 6.1 Gives personal information (social-emotional) =.121 6.2 Gives topic-related information =.248 6.3 Gives technical information =.327 Task Area: Questions < 0.1* 7 Asks for orientation, information, repetition, confirmation 7.1 Asks technical information =.30 7.2 Asks topic-related information <.05* 7.3 Asks personal information (social-emotional) =.01* 8 Asks for opinion, evaluation, analysis, expression of feeling =.07 9 Asks for suggestion, direction, possible ways of action =.07 Social-emotional Area: Negative Reactions <.0.5* 10 Disagrees, shows passive rejection, formality, withholds help =.44 11 Shows tension, asks for help, withdraws out of field <.05 12 shows antagonism, deflates other's status, defends or asserts self =.48 p. 4

In comparison of the two main areas in social-emotional oriented content and task-oriented content as shown in table 2, the mean sentence for social-emotional content in synchronous mode (68.18) is significantly higher than that in asynchronous mode (28.67). Task-oriented content consists of category 4, 5, 6.2, 6.3, 7.1, 7.2, 8, and 9. Social-emotional content consists of category 1, 2, 3, 6.1, 7.3, 10, 11, and 12. The breakdown in category 1, 2, and 3 shows that the mean sentence in synchronous mode (40.73, 5.36, 9.64 respectively) is much higher than those in asynchronous mode (7.5, 1.0, 6.5). In synchronous communication mode, there were more interactions in greeting, providing help, joking, and showing agreement. Whereas the mean sentence of task-oriented content in asynchronous mode (307.33) in significantly higher than that in synchronous mode (139.18). Student tended to give their opinions without asking other s opinions. It was done mostly through one-way communication in asynchronous discussion. The mean sentence of category 5 is significantly higher in asynchronous mode (247.33) than in synchronous mode (62.00). The interaction in task-oriented content indicated that in synchronous mode most of the discussions were two-way communication through the acts of providing and asking opinions on course-related subjects. Table 2: ANOVA table for Social-emotional oriented content vs. task-oriented content Asyn. mean Asyn. SD Syn. mean Syn. SD t-value F-value p-value SE content 28.67 20.81 68.18 39.95-2.24 5.016 0.041* TASK content 307.33 159.37 139.18 81.93 2.91 8.48 0.01** CONCLUSIONS Studying online interaction can further the understanding of how people communicate over the networks. The main difference between this study and other studies is the focus on the comparison of synchronous and asynchronous communication networks. The preliminary findings highlight the significant differences in both social-emotional oriented interaction and task-oriented interaction between the two communication modes. These differences indicate the very different nature of online interaction in both modes. Educators may want to design activities that will ensure a focused online discussion without diverting to only interpersonal exchanges in synchronous communication mode. Whereas in asynchronous mode, activity design to enhance interpersonal connection may facilitate more two-way communication. Further analysis on the effects of learner characteristics and learning tasks on interactions and the relationship between interaction and learning will be conducted in the near future. The research findings will contribute to the knowledge base of learner interaction in both synchronous and asynchronous networks. Educators, researchers, and instructional designers working on distance education can all make use of the knowledge to maximize the effectiveness of distance learning. ACKNOWLEDGEMENTS The author would like to acknowledge Xun Ge, Kelly Yamashiro, and Jenifer Winter for their assistance in coding the online transcripts and suggestions for this study. Also the author would like to thank Stephen Philion for his support and correction of this report. p. 5

BIBLIOGRAPHY Acker, S. R., & McCain, T. A. (1993). The contribution of interactivity and two-way video to successful distance learning applicatins: A literature review and strategic positioning, The Center for Advanced Study in Telecommunications. The Ohio State University, Columbus, Ohio. Bales, R. F. (1950). Interaction Process Analysis: A Method for the Study of Small Groups. Cambridge, MA: Addison-Wesley Press. Gunawardena, C. N., Lowe, C. A., & Anderson, T. (1998). Transcript analysis of computer-mediated conferences as a tool for testing constructivist and social-constructivist learning theories, Proceedings: Distance learning '98: The 14th Annual Conference on Distance Teaching and Learning (pp. 139-145). Madison, WI: University of Wisconsin. Hartman, K., Neuwirth, C. M., Cochran, C., Palmquist, M., & Zubrow, D. (1995). Patterns of Social Interaction and Learning to Write: Some Effects of Network Technologies. In M. Collins & Z. Berge (Eds.), Computermediated communication and the online classroom (Vol. 2, pp. 47-78). Cresskill, NJ: Hampton Press. Lewis, L., Snow, K., Farris, E., & Levin, D. (1999). Distance Education at Postsecondary Education Institutions: 1997-98, [Statistical Analysis Report NCES 2000-013). National Center for Education Statistics, U.S. Department of Education, Office of Educational Research and Improvement. Available: http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2000013 Martens, R. L., Valcke, M. M. A., Portier, S. J., Weges, H. G., & Poelmans, P. H. A. G. (1997). Research with interactive learning environments in three content domains: Descriptive statistics, continuous mathematics and substantive criminal law. Distance Education, 18(1), 44-58. McDonald, J., & Gibson, C. (1998). Interpersonal dynamics and group development in computer conferencing. The American Journal of Distance Education, 12(1), 7-25. Moore, M. G. (1989). Editorial: Three types of interaction. The American Journal of Distance Education, 3(2), 1-6. Newman, D. R., Webb, B., & Cochrane, C. (1995). A content analysis method to measure critical thinking in faceto-face and computer supported group learning. Interpersonal Computing and Technology: An Electronic Journal for the 21st Century, 3(2), 56-77. Paulsen, M. F. (1995). Moderating educational computer conferences. In Z. L. Berge & M. P. Collins (Eds.), Computer-mediated communication and the online classroom, volume three: distance learning (pp. 81-89). Cresskill, NJ: Hampton Press. Rice, R. E., & Love, G. (1987). Electronic Emotion: Socioemotinal Content in a Computer-Mediated Communication Network. Communication Research, 14(1), 85-108. Rohfeld, R. W., & Hiemstra, R. (1995). Moderating discussions in the electronic classroom. In Z. Berge & M. Collins (Eds.), Computer-mediated communication and the online classroom, volume three: distance learning (pp. 91-104). Creskill, NJ: Hampton Press. Scardamalia, M., Bereiter, C., Brett, C., Burtis, P. J., Calhoun, C., & Lea, N. S. (1992). Educational applications of a networked communal database. Interactive Learning Environments, 2(1), 45-71. Short, J., Williams, E., & Christie, B. (1976). The social psychology of telecommunications. New York, NY: John Wiley. Sorensen, G., & McCroskey, J. C. (1977). The prediction of interaction behavior in small groups: Zero history vs. intact groups. Communication Monographs, 44, 73-80. Wagner, E. D. (1998). Interaction strategies for online training designs. Proceedings of Distance Learning 98: 14th Annual Conference on Distance Teaching & Learning (pp. 417-421). Madison, WI: University of Wisconsin System. p. 6