Rigorous interaction between peers has been an elusive goal in online asynchronous discussions. Intersubjectivity, the goal of peer-to-peer interaction, is a representation of a higher quality of synthesis. It is the representation of knowledge construction achieved through a synergistic progression from individual contributions to sequences of interdependent contributions. The content analysis in this research examined the intersubjectivity of peer responses of 2 semesters of a practical online educational technology course and 1 semester of a theoretical worldview course. The peer responses were analyzed using the interaction analysis model. Results found higher levels of intersubjectivity in the theoretical course than the practical courses. It may be more difficult to achieve intersubjectivity in practical, skills-based courses than in theoretical courses, even when knowledge construction is the intended outcome of both types of courses.
Introduction
In the past decade, online education has moved from the margin to mainstream with many institutions offering online programs and courses. A great deal of variety exists in the kinds of online learning experiences available; however, the discussion forum remains a staple of the online learning experience. Regular interaction with the instructor is a required component to qualify as a distance education course, as opposed to a correspondence course, according to the United States federal definitions (Office ofthe Federal Register, 2010). Some accrediting bodies also require a learning communityamong the students (Association of Theological Schools, 2010; Higher Learning Commission, 2009). Even so, rigorous interaction between peers remains an elusive goal in asynchronous discussions within onlin courses (Dooley & Wickersham, 2007; Hall, 2010; J. R. Morrison, Watson, & Morrison, 2012).
One emerging notion to address the need for rigorous interaction is the concept of intersubjectivity. Intersubjectivity represents the higher quality of synthesis necessary to achieve the knowledge construction essential in a social constructivist environment (Vygotsky, 1978). Knowledge construction is not a product of students just posting messages in the discussion board; instead, negotiation and action-reaction sequences are required to lead toward intersubjectivity (Dennen & Wieland, 2007). Intersubjectivity is “the representation of knowledge construction achieved through a synergistic progression from individual contributions to sequences of interdependent contributions” (Hall, 2011, p. 25). Martin, Sokol, and Elfers (2008) defined intersubjectivity as the ability to take other perspectives and reflectively integrate them ininteractions with others. Suthers (2006) proposed that intersubjectivity may occur incomputer supported collaborative learning when students jointly create interpretationsthrough participation, dissonance, and interaction to establish common ground. Dialogue online usually takes the form of threaded discussions, “a hierarchically organized collection of notes in which all notes but one (the note that started the thread) are written as ‘replies’ to earlier notes” (Hewitt, 2005, p. 568). Peer responses representing conversational turn taking interwoven together may stimulate higher levels of critical thinking and knowledge construction.
The concept of intersubjectivity raises several questions. How can intersubjectivity help instructors achieve higher levels of critical thinking within the social construction of knowledge occurring in discussion forums? Do some content areas lend themselves more to intersubjectivity than others? This article proposes that some course content, such as theoretical studies, may be more likely to stimulate higher levels of intersubjectivity, while other content, such as skills-based courses, may find most learning occurring at lower levels of intersubjectivity.
Intersubjectivity in Social Constructivist Online Discussions
Social Constructivism in Online Discussion
Social constructivism is currently held in esteem within the online learning world.The theory suggests we make meaning within social contexts. Social experiences provideopportunities to share our divergent views of the external world and to construct our knowledge as we interact with others (Bruner, 1987; Jonassen, 1991; Vygotsky, 1978). Many online courses are designed with the intent to use the discussion forum to facilitate the social construction of knowledge, including the courses examined in this study. Despite the importance of peer-to-peer discussions within social constructivist learning environments, more than two decades of research (see Hall, 2010, 2011) consistently demonstrated that asynchronous discussions could offer more substantial benefit toward course learning outcomes.
While some faculty who teach discussionrich seminar classes on campus may be hesitant to use the online format (K. A. Morrison, 2011), social constructivist discussion forums are used for different learning outcomes. Freed (2003) maintained the value of the discussion forum as promoting reflective dialogue, providing an opportunity for participants to try out new ideas in the context of their experience and the experiences of the other class participants. Outcomes for learners in online constructivist environments include becoming self-directed, learning to integrate multiple perspectives, and evaluating different arguments (Tam, 2000). Online discussions motivate students, engage students with real-world, task-oriented activities, and help students develop their voice (Milman, Hillarious, & Walker, 2012; Rovai, 2007). Other outcomes include developing critical thinking (Greenlaw & DeLoach, 2003; Oncu & Cakir, 2011), social awareness (Ligorio, Cesareni, & Schwartz, 2008), democratic discussion (Armstrong & Thornton, 2012) and collaborative learning (Thompson & Heng-Yu, 2006).
Interventions to Improve Quality of Online Discussions
Discussion forums have been a staple of online learning for decades, and many interventions have been researched to improve the quality of online discussions. J. R. Morrison et al. (2012) highlighted the conflicting research on the level of instructor involvement required for a quality interactive experience. Rovai (2007) contended that instructors must find a balance within the discussion—attending to social equity issues while avoiding dominating the discussion. Instructors are critical to the success of an online discussion (Blignaut & Trollip, 2003; Greenlaw & DeLoach, 2003; Jackson, 2010; Roblyer & Wiencke, 2003) as a “learning catalyst and knowledge navigator” (Volery, 2001, p. 77). The instructor’s role requires facilitation skills—probing, redirecting, asking questions, correcting misunderstandings, and supporting student-student interactions (Bober & Dennen, 2001; Curry & Cook, 2014; Rovai, 2007). Students expect the instructor to provide relevant feedback and consistent just-in-time support (Du & Xu, 2010; Fayer, 2014). Instructors promote quality discussion by providing instructions on creating and evaluating arguments (Jeong, 2004; Poscente & Fahy, 2003).
Teaching presence, a well-known instructional intervention identified by Anderson, Rourke, Garrison, and Archer (2001), includes multiple indicators within the three categories of direct instruction, facilitating discourse, and instructional design. The instructor must provide a clear framework for grading (Dixon, 2014; Eccarius, 2011; Langille & Pelletier, 2003) and administrative guidelines (Baran, Correia, & Thompson, 2011). Spatariu and Winsor (2013) reiterated the necessity of a discussion grading rubric with expectations on writing, tone, and peer collaboration. Indeed, some students lack substantive participation in discussion due to the lack of standardguidelines (Dennen, Darabi, & Smith, 2007; Farmer, 2004) and concerns of not meeting the instructor expectations (Han & Crooks, 2013). Communication of strategies facilitates students developing requisite knowledge (Sherblom, Withers, & Leonard, 2013).
Some have proposed frameworks for promoting and evaluating interaction and criticalthinking in online discussion. Curry and Cook (2014) recommended students create discussion posts using the acronym MANIC, by sharing the Most important thing in the reading, something they Agreed with, something they did Not agree with, something Interesting, and something Confusing. Dixon (2014) created a simple model of discussion organized around three E’s: the Experience of creating an online community, using Engagement to teach students to be active learners, and Evaluation with clear objectives and fair evaluation by the instructor. Freed (2003) proposed encouraging students to usemetaphors to extend reflection and dialogue. The interaction analysis model encouragesparticipants to find areas of disagreement, followed by identifying new constructions of knowledge and proposing synthesis of multiple perspectives (Gunawardena, Lowe, & Anderson, 1997). Marra, Moore, and Klimczak (2004) also suggested using a five stage critical thinking model moving through identification of the problem, definition of the problem, followed by exploration, evaluation, and integration (Garrison, 1992; Newman, Johnson, Webb, & Cochrane, 1997; Newman, Webb, & Cochrane, 1995). J. R. Morrison et al. (2012) built on Kanuka’s (2005) work to classify online postings as prestructural (the student’s response is not appropriate), unistructural (the student addresses one aspect of the discussion prompt), multistructural (each concept is treated separately), relational (the aspects are integrated into a coherent whole), and extended abstract (includes metacognition). Thompson and Heng-Yu (2006) measured interdependence of group participation by examining patterns of interaction, such as a-b-c-x where participants a, b, and c interacted and created a solution (x) at the end of the interaction. The discussion analysis tool can be used to analyzecritical thinking events (Jeong, 2003). Each model targets synthesis, higher levels ofcritical thinking, and interdependence between the participants in the online discussion forum.
Intersubjectivity in Online Discussions
Another concept aiming for higher levels of critical thinking is intersubjectivity. Intersubjectivity moves constructivist interaction to a higher level through a synergistic progression from individual contributions to sequences of interdependent contributions (Dennen & Wieland, 2007; Martin et al., 2008). Intersubjectivity involves learner interdependence, synergistic communication, and participation progressing toward knowledge construction. The definition of intersubjectivity is drawn from the definitions of other researchers in the fields of sociology, psychology, and philosophy. According to Matusov (1996), intersubjectivity represents how individual contributions are coordinated with each other during the activity, thereby creating “continuity in activity progression” through “building on each other’s contributions” (p. 41). Matusov’s description of intersubjectivity is consistent with the description of Baker, Hansen, Joiner, and Traum (1998) as the “coordination of contributions in joint activity … not just overlapping of conceptualizations” (p. 4). For Bober and Dennen (2001), intersubjectivity is the development of shared understanding, relying on ongoing conversation artifacts to develop new contributions to the discourse. Intersubjectivity is building something new from the combination of different perspectives (Ligorio et al., 2008; Martin et al., 2008).
Theoretical Versus Practical Course Content
While intersubjectivity may be helpful in increasing the levels of critical thinking within online discussion, is there a distinction between different types of course content? Some research has compared blended, online, and face-to-face discussion (Benbunan-Fich & Hiltz, 2003; Braun, 2008; Rovai & Jordan, 2004) or class size (Hewitt & Brett, 2007). Andresen (2009) shared the challenges of problem solving for math and science in asynchronous discussions, how forums may ease the discussion sensitive multicultural issues, and how students may feel disconnected from the discussion. Hattinger, Spante, and Ruijan (2014) maintained that theoretical engineering courses should have more interaction among the instructor and students, while practical courses should include learning activities teaching design and materials selection. Apostolopoulos (2014) argued that in a skills-based course, the instructor-student interaction is more important than student-student interaction, while Billett (2001) emphasized intersubjectivity developing within the work place where practical knowledge is constructed. However, very little research has been conducted on the influence of course content on the effectiveness of online discussion.
Purpose of the Study
While discussion forums have been researched often, the concept of intersubjectivity as a method for promoting critical thinking deserves further attention. Oncu and Cakir (2011) proposed four main research priorities for online education, including learner engagement, which can be promoted through intersubjectivity in a social constructivist learning environment. Promoting intersubjectivity within an online discussion is a delicate task for the designer and instructor, and the orchestration of learning requires additional research (Lucas, Gunawardena, & Moreira, 2014). This content analysis examined the intersubjectivity of peer responses of two semesters of a practical online educational technology course and one semester of a theoretical worldview course. While both courses included application of the knowledge, the educational technology courses focused on skills-based knowledge; while the worldview course focused on theoretical knowledge. The purpose of this study was to examine the variance of knowledge construction in skills-based content knowledge vs theoretical content knowledge measured by intersubjectivity within peer responses. The research question was: How does the theoretical or practical nature of the course content influence the intersubjectivity within peer responses?
Methods
Research Context
Data presented were collected from 6 weeks each of two semesters of an online educational technology course for teachers and 6 weeks of one semester of an online worldview course. Two semesters of the practical online educational technology course were chosen because of the low number of participants in each course. All three graduate courses were offered for two semester credits through a private university in the Midwest attracting participants from all over the United States, Canada, and internationally. Two different instructors taught the courses; however, both instructors designed the courses from a social constructivist perspective, expecting a significant amount of learning to occur from interaction with peers. The instructors’ educational philosophies deeply value the interaction between students, the importance of choice for adult learners, and the prior experience adults bring to learning tasks.
The practical course addressed how to integrate technology in the classroom with a faithbased K–8 integrated literacy curriculum. The course readings focused on using technology to enhance thematic instruction, vocabulary instruction, concept mapping, guided reading, information literacy, and assessment. Participants read three selected articles each week and reflected on their practice in the context of the readings, discussing potential new teaching practices in a weekly forum. The course application weekly assignments addressed four themes of the literacy curriculum and featured several technology tools such as concept mapping software, free online slideshow tools, Google Earth, Skype, social bookmarking, and wikis to design projects to support the integrated literacy curriculum. Participants created a project example to share in another weekly forum. Project choices each week provided significant opportunities for participants to learn from each other’s diverse approaches to the learning task. Some participants during the fall session actually completed the weekly projects with their students, applying learning immediately in the classroom throughout the course. As a final project, each participant designed a thematic unit or a learning center plan to supplement the Pathways curriculum with technology-supported learning experiences. The instructor’s role within the discussion forum included sharing resources, answering questions, providing technical assistance on the various technology tools, and asking questions to encourage deeper thinking.
The discussion requirements of the practical course were improved in the second section. In the reading forums in the first session of the course, participants were encouraged to respond to peers giving feedback, sharing their experiences, asking questions, building on previous peer responses, and suggesting metaphors. Participants were required to reply to at least two classmates. In the weekly application forum, participants were encouraged to share resources, similar experiences, feedback, and questions. Participants were required to give feedback to at least one peer each week. The second session of the course was revised with a peer response rubric added and required replies increased to a minimum of four peer responses to classmates in each weekly forum. Participants were encouraged to participate at least three times per week.
The theoretical course addressed the purpose of education, the nature of knowledge, diverse worldviews, and the program’s philosophical tenants. Participants were required to articulate their own worldview and make connections between worldviews and leadership. The course was organized into five “chunks” of 3 weeks each. Each chunk included 2 weeks of significant reading and discussion, followed by a quiet week for individual reflection on their own worldview. The 6 weeks of discussion data researched covered the first three chunks of the course: purpose of education and nature of knowledge; approaches to learning about worldview; and applying my worldview. The course included a weekly required post within a private worldview forum to prepare for writing the final paper; a role play group project; sharing resources in a course library; and the final philosophical foundations paper. The instructor’s role within the discussion forum included asking questions, probing thinking, sharing additional resources, making connections to required reading, and encouraging student reflection.
The discussion forum instructions required participants to write three to four substantive posts each week. No explicit differentiation was made between replies to the instructor’s initial question or peer responses in the points given for the discussion forum; however the instructor expectations required giving evidence of listening, asking for clarification, sharing experiences, affirming others, and extending the conversation. Participants were encouraged to create meaningful posts by building on previous posts, agreeing or disagreeing, giving examples, and defining issues. Participants received full points for sharing during the weekly time frame, and posting early and later in the week.
Participants
The participants in the practical courses were multigrade teachers in private elementary schools in Canada and the United States. In the first section, six participants taught Grades K– 4 and three taught Grades 5–8. Seven were female and two were male. In the second section, two participants taught Grades K–4, two taught Grades 3–6, and two taught Grades 5–6. Five of the participants were female and one was male. The participants and their schools were all part of the same network of private schools.
The 20 participants in the theoretical course were enrolled in master’s and PhD programs in leadership from Canada, Germany, Hong Kong, and the United States. The participants held diverse leadership roles: business owner, communication director, hospital administrator, K12 teacher, pastor, registrar, social worker, treasurer, and university administrator. Nine of the participants were female and eleven were male.
Research Design
The research design for this study was a descriptive quantitative content analysis. Content analysis is a systematic analysis of text, in this case the text generated in the online discussion forum (Neuendorf, 2002; Rourke & Anderson, 2004). Content analysis may be used to analyze words, phrases, sentences, or paragraphs; however, in this study, each message unit was selected for analysis. Message units are determined by the author of the message, easily identifiable, and produce a convenient number of cases (Rourke, Anderson, Garrison, & Archer, 2001).
This research examined peer responses manifested as discussion forum replies. Instructor messages and responses to the instructor were not included. The restriction to peer responses was chosen to study what Suthers calls “uptake acts in which one participant takes up another’s contribution and does something further with it” (2006, p. 331), also known as intertwined action-reaction sequences (Dennen & Wieland, 2007; Walther, Gay, & Hancock, 2005).
Measures
Several measures were used to understand the peer response quality. The number of words measured the peer response length. The number of citations was measured by counting the number of unique references to literature, peers, websites, famous quotations, or online sources within the peer response.
The interaction analysis model (IAM) was selected to measure intersubjectivity within the discussion threads of the online course rooms. The IAM has been studied more than any other discussion analysis model in the published literature (Gunawardena et al., 1997; Hall, 2011, 2015). The IAM was selected as an a priori coding scheme to determine the categories for the content analysis (Marra et al., 2004). Using existing content analysis protocols, rather than creating new ones, adds to overall validity of the existing protocol (Rourke & Anderson, 2004). The IAM has high levels of inter-rater reliability among the published studies where reliability is reported (Lucas et al., 2014).
The IAM has five phases learners experience as knowledge is being constructed: (1) sharing and comparing; (2) inconsistency and dissonance; (3) negotiation of meaning and coconstruction of knowledge; (4) testing and modification of proposed co-construction; and (5) applications of newly constructed meaning or summarizations of agreement. The IAM model measures the knowledge being constructed. Garrison, Cleveland-Innes, Koole, and Kappelman (2006) argue that a reliable coding scheme must have meaningful categories, discernable indicators, and manageable message units. Each phase of the IAM model includes descriptions of discernable indicators to assist the coder in determining which phase(s) are evidence in the peer response. When analyzing each post, the most advanced phase was applied to the whole peer response, as suggested by previous researchers studying online discussion boards (Beaudrie, 2000; Gunawardena et al., 1997; Marra et al., 2004). Consistent with previous studies (Beaudrie, 2000; Luebeck & Bice, 2005; Moore & Marra, 2005; Onrubia & Engel, 2009), each post was coded according to the highest phase of the IAM present within the response. Thus, a post demonstrating several indicators of different phases was assigned a single code according to the highest phase represented in the response.
Procedures
All of the discussion posts from the three courses were collected and analyzed to identify the peer responses (Dennen & Wieland, 2007; Suthers, 2006). The peer responses were defined as discussion posts replying to other course participants. Social posts were not included in the data set (Gunawardena et al., 1997; Henri, 1992).
Four researchers, including the two authors, analyzed the data for the IAM phase. One researcher analyzed the practical courses; the second researcher analyzed the theoretical course. The second author provided coding training; and both authors participated in interrater agreement analysis.
Training was provided to each researcher on how to code for the IAM phase. Following the coding procedure described by Marra et al. (2004), each discussion thread in all three courses was read sequentially, and each post (n = 1,328) carefully examined to determine the existing phase. Interrater differences were dealt with using a negotiated protocol based on IAM research (Garrison et al., 2006; Gunawardena et al., 1997; Marra et al., 2004). With the first two practical courses, the researcher independently analyzed all of the peer responses for the IAM phase. Then the two authors independently analyzed a random 10% of the peer responses. The interrater agreement between the authors and the first researcher was 93% and 91% respectively. As noted in the results section, the vast majority of the posts in the practical courses were at Phase 1, making interrater agreement easier to achieve. In the theoretical course, after the fourth researcher coded all of the peer responses for the IAM phase, the two authors independently analyzed a random 10% of the peer responses. The initial inter-rater agreement between the authors and the fourth researcher was 45% each. The researchers then discussed the discrepancies and worked independently to code another 10 posts. The researchers then achieved 70% interrater agreement, and in discussion of the second set of discrepancies, the researchers came to 100% agreement on the coding. The fourth researcher then recoded all of the peer responses. This negotiated method is recommended for content analysis where not all of the researchers are familiar with the coding rubric (Garrison et al., 2006).
Results
This section reports a description of the peer responses in the three courses and the results related to the intersubjectivity of the peer responses.
Peer Responses Descriptive Statistics
This research analyzed the peer responses within the 6 weeks selected from the three online courses. The descriptive statistics of the peer responses are shown in Table 1. The first course, Course A, had 352 peer responses written by nine participants. Course B had 487 peer responses written by six participants.
Peer Response Descriptives for Each Course
| n | |||||
|---|---|---|---|---|---|
| Total Posts Per Course | |||||
| Course A with 9 participants | 468 | ||||
| Course B with 6 participants | 588 | ||||
| Course C with 20 participants | 614 | ||||
| Number of Peer Responses Per Course | |||||
| Course A with 9 participants | 352 | ||||
| Course B with 6 participants | 506 | ||||
| Course C with 20 participants | 351 | ||||
| Average Number of Peer Responses Per Participant | |||||
| Course A | 39.11 | ||||
| Course B | 103.67 | ||||
| Course C | 17.55 | ||||
| Average Number of Total Posts Per Week Per Participant | |||||
| Course A | 8.67 | ||||
| Course B | 16.30 | ||||
| Course C | 5.12 | ||||
| Average Number of Peer Responses Per Week Per Participant | |||||
| Course A | 6.52 | ||||
| Course B | 14.06 | ||||
| Course C | 2.93 | ||||
| n | Min | Max | Mean | SD | |
| Average Number of Words Per Peer Response | |||||
| Course A | 352 | 2 | 572 | 70.97 | 71.738 |
| Course B | 506 | 4 | 533 | 91.62 | 63.122 |
| Course C | 351 | 3 | 982 | 204.84 | 156.520 |
| Average Number of Citations Per Peer Response | |||||
| Course A | 352 | 0 | 1 | .01 | .092 |
| Course B | 506 | 0 | 0 | .00 | .000 |
| Course C | 351 | 0 | 5 | .71 | .879 |
| n | |||||
|---|---|---|---|---|---|
| Total Posts Per Course | |||||
| Course A with 9 participants | 468 | ||||
| Course B with 6 participants | 588 | ||||
| Course C with 20 participants | 614 | ||||
| Number of Peer Responses Per Course | |||||
| Course A with 9 participants | 352 | ||||
| Course B with 6 participants | 506 | ||||
| Course C with 20 participants | 351 | ||||
| Average Number of Peer Responses Per Participant | |||||
| Course A | 39.11 | ||||
| Course B | 103.67 | ||||
| Course C | 17.55 | ||||
| Average Number of Total Posts Per Week Per Participant | |||||
| Course A | 8.67 | ||||
| Course B | 16.30 | ||||
| Course C | 5.12 | ||||
| Average Number of Peer Responses Per Week Per Participant | |||||
| Course A | 6.52 | ||||
| Course B | 14.06 | ||||
| Course C | 2.93 | ||||
| n | Min | Max | Mean | SD | |
| Average Number of Words Per Peer Response | |||||
| Course A | 352 | 2 | 572 | 70.97 | 71.738 |
| Course B | 506 | 4 | 533 | 91.62 | 63.122 |
| Course C | 351 | 3 | 982 | 204.84 | 156.520 |
| Average Number of Citations Per Peer Response | |||||
| Course A | 352 | 0 | 1 | .01 | .092 |
| Course B | 506 | 0 | 0 | .00 | .000 |
| Course C | 351 | 0 | 5 | .71 | .879 |
Course C had 351 peer responses written by 20 participants. Course B had the highest average number of peer responses per participant (Course A = 39.11; Course B = 103.67; Course C = 17.55) and the highest average peer responses per participant per week (Course A = 6.52; Course B = 14.06; Course C = 2.93). Participants in Course C wrote the peer responses with the highest average number of words per peer response (204.84). Participants in Course C wrote peer responses with the highest average number of citations per peer response (0.71).
In Table 2, descriptive statistics are shown for the interaction analysis model phases measuring intersubjectivity. Course A and B con
Interaction Analysis Model Phases
| Phase I | Phase II | Phase III | Phase IV | Phase V | |
|---|---|---|---|---|---|
| Course A | 98% | 2% | |||
| Course B | 100% | ||||
| Course C | 70.4% | 18.2% | 9.1% | 0.6% | 1.7% |
| Phase I | Phase II | Phase III | Phase IV | Phase V | |
|---|---|---|---|---|---|
| Course A | 98% | 2% | |||
| Course B | 100% | ||||
| Course C | 70.4% | 18.2% | 9.1% | 0.6% | 1.7% |
sisted primarily of peer responses at the Phase 1 of sharing and comparing. The peer responses in Course C were also predominately at Phase 1; however, all of the other IAM phases are also represented in Course C’s peer responses.
Table 3 outlines the discussion expectations and intersubjectivity across the three courses. Courses A and B required an initial post responding to the instructor’s question and additional peer responses. Course C did not differentiate between initial posts and peer responses. Course B and C included a rubric for the peer responses and the participants created longer posts. Course C had a higher average IAM phase, however including a rubric in Course B did not result in a higher IAM phase. Results are mixed, with Course C resulting in longer posts and a higher average IAM phase.
Discussion Expectations Effect on Peer Responses
| Number of Initial Posts Required Per Week | Number of Peer Responses Required Per Week | Rubric Included for Peer Responses | Word Length Requirement | Average IAM Phase | Average Length in Words | |
|---|---|---|---|---|---|---|
| Course A | 2 | 3 | No | None | 1.02 | 70.97 |
| Course B | 2 | 8 | Yes | None | 1.00 | 91.62 |
| Course C | combined with peer | 3−4 total posts | Yes | None | 1.45 | 204.84 |
| Number of Initial Posts Required Per Week | Number of Peer Responses Required Per Week | Rubric Included for Peer Responses | Word Length Requirement | Average IAM Phase | Average Length in Words | |
|---|---|---|---|---|---|---|
| Course A | 2 | 3 | No | None | 1.02 | 70.97 |
| Course B | 2 | 8 | Yes | None | 1.00 | 91.62 |
| Course C | combined with peer | 3−4 total posts | Yes | None | 1.45 | 204.84 |
Intersubjectivity of Peer Responses
The research question examined the difference between the IAM phase for the practical courses (Course A and B) and the theoretical course (Course C) and results are shown in Table 4. A Mann-Whitney U test was run to rank and then identify any significant differences in the IAM Phase of the peer responses between the practical courses and the theoretical course. Distributions of the IAM phase for the practical and theoretical courses were not similar, as assessed by visual inspection. The IAM phase scores for the peer responses in the theoretical course (mean rank = 729.01) were statistically significantly higher (U = 194106.5, z = 15.776, p < .001) than the IAM phase scores for the peer responses in the practical courses (mean rank = 554.27).
Intersubjectivity of Peer Responses
| Group | n | Mean Rank | U | z | P | |
|---|---|---|---|---|---|---|
| IAM Phase | Practical Courses | 858 | 554.27 | |||
| Theoretical Course | 351 | 729.01 | ||||
| 194106.5 | 15.776 | .001 |
| Group | n | Mean Rank | U | z | P | |
|---|---|---|---|---|---|---|
| IAM Phase | Practical Courses | 858 | 554.27 | |||
| Theoretical Course | 351 | 729.01 | ||||
| 194106.5 | 15.776 | .001 |
In summary, this study examined the intersubjectivity of peer responses in two practical courses and one theoretical course. The discussion expectations between the courses did not make a difference between the courses; however the theoretical course had longer posts and more citations. Peer responses within the theoretical course achieved higher levels of intersubjectivity as measured by the IAM phase than the peer responses in the practical course.
Discussion
Understanding the differences in how theoretical courses and practical courses reach higher levels of intersubjectivity has implications for research and faculty support. Previous research has established the challenges of designing discussion forums where students achieve higher levels of intersubjectivity (Garrison, Anderson, & Archer, 2001; Gunawardena et al., 1997; Lucas et al., 2014; J. R. Morrison et al., 2012; Rourke & Kanuka, 2009). However, limited research has explored the influence of the nature of the course content on discussion. Practical courses may not generate the disagreement necessary to move to phase three and higher within the IAM model. On the other hand, discord may not be the only factor that triggers higher levels of intersubjectivity. Student-created metaphors may trigger deeper thinking (Freed, 2003). When considering how students build knowledge across time, the concept of fantasy chaining may provide further insights to the facilitation of online discussions (Hirokawa & Salazar, 1999).
The actual design of the discussion forum may also contribute to higher levels of intersubjectivity. Hewitt (2005) argued threads die because participants only read the discussion forum once, and when no more new posts are written within a thread, students and instructor stop reading that thread. Jeong (2004) maintained that response time to threads may also be a factor. Within the theoretical course, participants considered a concept over 3 weeks. The first 2 weeks they discussed the concept in one discussion forum; and in the third “silent week,” they continued working in their personal worldview space. Participants may have synthesized more because of a deep engagement in the content trigged by the design of the course. Mayo (2003) found increased critical thinking within a course where students wrote in an observational diary, similar to the personal worldview space in the theoretical course. The limited number of posts per participant and the long length of the participant’s posts was another course component unique to the theoretical course. Darabi and Jin (2013) suggested that limiting the number of posts may reduce the challenges of cognitive overload in heavy online discussions. Writing fewer yet longer posts may be more conducive to synthesis and critical thinking. Finally, a third course design difference was the organization of peer responses and responses to the instructor. In the practical courses, a clear line defined an initial response to the instructor’s question and a peer response. The theoretical course’s discussion was more free-flowing. Students asked starter questions and the instructor played the role of a peer. Evaluating only responses to peers analyzed about half of the total posts in the theoretical course. Additional clues to the higher levels of intersubjectivity may be found in other posts where participants may have summarized thinking over several weeks of the course or responded to the instructor instead of a peer. Instructor’s questions within the discussion may also contain insights to the higher level of knowledge construction. Further research could analyze the implications of analyzing only peer responses as opposed to analyzing the full discussion including the instructor’s role.
The design of the courses studied included several research-supported components necessary for high quality discussions. For example, the utilization of a grading framework can improve the quality of online discussions (Dixon, 2014; Langille & Pelletier, 2003; Spatariu & Winsor, 2013). In this study, the grading rubric for the second practical course was created from the Interactive Analysis Model; while the grading rubric in the theoretical course was influenced by work emphasizing dialogue over argument (Freed, 2003; Isaacs, 1999; Tannen, 1999). Other scaffolding included instructor created structure (Lai, 2015), such as a note starter to help students formulate critical thinking responses (Nussbaum, Hartley, Sinatra, Reynolds, & Bendixen, 2002; Spatariu & Winsor, 2013), student questions as discussion triggers (Poscente & Fahy, 2003), and interesting discussion topics (Han & Crooks, 2013). Results suggest theoretical course content may be more conducive to reaching higher levels of critical thinking and synthesis than practical courses.
The interaction analysis model was created in the study of a strictly structured debate, accounting for the second phase of disagreement within the model (Gunawardena et al., 1997). The theoretical course content inherently had debate, disagreement, and dissonance as the participants considered different worldviews, while the practical course content focused on application of technology tools with limited opportunity for dissonance. The higher levels of the IAM may be challenging for the construction of knowledge without much dissonance. Even Gunawardena et al. (1997) admitted the debate format is not conducive to the co-construction of knowledge because participants wanted to synthesize, but the debate facilitators tried to keep participants on their side of the debate. In addition, women may be less likely to disagree or confront peers (Jeong & Joung, 2003) and may prefer dialogue over taking opposing sides (Tannen, 1999). In later research, Jeong (2007) found a significant difference in the number of rebuttals posted between men and women. The higher percentage of women in the practical courses could possibly explain the low levels of disagreement. Further research could explore the influence of gender and debate format on levels of intersubjectivity in online courses.
Gunawardena et al. (1997) proposed two types of learning: (1) pooling knowledge and elaboration and (2) creating new meaning where participants must alter their thinking to integrate new concepts with preexisting cognitive schema. The theoretical class provided opportunities for adjusting thinking; while the practical courses provided more opportunities for pooling knowledge. Further research on the type of course content could explain the persistence of low IAM phases in online discussion research.
Others have suggested metaphors may trigger higher levels of critical thinking (Freed, 2003). Students may explore ideas in the form of a metaphor, without giving textual evidence of conflicted thinking within the discussion forum. The indicators of the second phase of the IAM do include the “proposal of a relevant metaphor or analogy,” but within the context of dissonance (Gunawardena et al., 1997, p. 414). Some students may not yet be at a thinking stage to integrate metaphors with course concepts. Further research could explore whether the participant’s ways of knowing (Belenky, 1997; Perry, 1970) are a factor in their ability to reach higher levels of intersubjectivity. A better understanding of what triggers students moving from sharing to higher levels of synthesis is needed.
Conclusion
This study found higher levels of intersubjectivity as measured by the IAM phase within the theoretical course than the practical courses. The results suggest the type of content and learning within the course may be an important factor to consider when researching the level of critical thinking, metacognition, and intersubjectivity of peer responses in online discussion forums. Instructors aiming to reach higher levels of intersubjectivity should consider designing discussion prompts inviting dissonance, trying on new ideas, creating metaphors to explore concepts further. Participants should be taught how to disagree appropriately; instructors should model how to synthesize the varying viewpoints shared in the peer responses. Instructors of practical courses desiring higher levels of intersubjectivity should carefully review the content to find the areas of theoretical concepts where students could consider dissonance, create metaphors, and synthesize new constructions of knowledge.
Reaching high levels of intersubjectivity is desirable, yet difficult in online courses (Bober & Dennen, 2001; Hall, 2010; Martin et al., 2008; Pawan, Paulus, Yalcin, & Chang, 2003; Suthers, 2006). Some questions suggest further research into the course design. Do some types of discussion prompts (Spatariu & Winsor, 2013) generate higher levels of intersubjectivity? Are there critical instructor behaviors or course design components for reaching higher levels of intersubjectivity? Future research could clarify factors such as instructor behavior, the content of the course, and the course design.
