Elsevier

Computers & Education

Volume 60, Issue 1, January 2013, Pages 172-183
Computers & Education

Subject line preferences and other factors contributing to coherence and interaction in student discussion forums

https://doi.org/10.1016/j.compedu.2012.07.005Get rights and content

Abstract

A number of factors may affect student interaction in an asynchronous online discussion forum used in learning. This study deals with student preferences for the subject line of messages and in what ways the choice of subject line contributes to coherence and interaction reflected in the textual and interpersonal functions of the linguistic items used. The study also attempts to determine what affects the choices made by participants. Nine separate discussion forums from three different undergraduate courses in English at a Swedish university were used in the study. A total of 98 students and 435 student messages were examined and a number of trends appeared. The functions of the subject line may be summarized as contributing to coherence by reflecting message content in a number of different ways. In addition, the subject line can perform other tasks such as maintaining social relationships among the participants. It is not clear in what ways the subject line contributes to interaction with regard to increasing the reading rate. The trends observed indicate that other factors than subject line content may contribute to whether students are inclined to access a message or not, such as when a message is posted and where it is displayed on the screen.

Highlights

► Functions of subject lines – discourse-organizing, stance- revealing and social. ► Group preferences for types of subject lines were observed. ► Subject line type did not correlate to reading rate.

Introduction

A large portion of studies at tertiary level takes place in a computer-mediated environment. For some students, participating in an online learning environment can be less stressful than the traditional classroom. A number of studies have shown that students who are less active in face-to-face interaction are more active online (Chun, 1994; Kelm, 1992; Kern, 1995). Some of the advantages experienced by students are that participants in asynchronous text-based discussions are able to edit their texts before publishing and they are also able to trace back through the whole discussion from its beginning (Hammond, 2000: 255). However, coupled with the potential advantages of the asynchronous online discussion forum are the challenges posed. One such challenge for instructors is getting students to interact with one another by reading and responding to the messages posted in an appropriate and efficient way. Merely providing an opportunity for interaction does not mean it will automatically happen. Learning more about the function of linguistic items used in the subject line may benefit both instructors and software developers so that they are better able to evaluate the way courses as well as software are designed for asynchronous discussions.

Many studies have examined different aspects of asynchronous discussion forums, among other things indicators of cognitive activity (Garrison, Anderson, & Archer, 2000; Henri, 1991; Schellens & Valcke, 2006), social presence and community building (Gunawardena & Zittle, 1997; Haythornthwaite, Kazmer, Robins, & Shoemaker, 2000; Richardson & Swan, 2003; Rourke, Anderson, Garrison, & Archer, 1999; Rovai, 2002) and influence of the instructor (Mazzolini & Maddison, 2007). Other studies have attempted to determine to what extent learning and perceived learning takes place in these communities (see Rourke & Kanuka, 2009 for a summary). As a result, models of and suggestions for how to improve interaction in asynchronous discussion forums have been put forth (Akin & Neal, 2007; Gilbert & Dabbagh, 2005). One aspect which has been neglected in previous studies is the role of the subject line; hence this study attempts to contribute to the already existing research on discussion forums by focussing on the function of the subject line and its potential to affect coherence and interaction.

The present study is concerned with communication in online asynchronous discussion forums situated in the learning management system (LMS) Fronter.1 By examining the linguistic items participants choose, the study attempts to shed light on the role of the subject line content of messages in the discussion forum interaction. The questions which it is hoped can be answered, are firstly in what ways linguistic items in the subject line contribute to overall coherence and whether there are other functions performed by these items. Secondly, it attempts to determine what preferences the students have for subject line type, and thirdly to what extent the subject line content as opposed to other factors affects whether other students access a particular message. Both course instructors and software developers present discussion forum instructions that contain recommendations for what constitutes an appropriate subject line.2 The purpose may seem obvious, yet there seems to be little research on the function and effect of the subject line on interaction in asynchronous discussion forums and to what extent participants use it the way it is intended. It is hoped that the present study will contribute to both instructors' and students' understanding of the choices made by participants in asynchronous text-based discussion forums for learning and how this knowledge might be used to enable students to interact more efficiently.

Most research on asynchronous discussion forums and asynchronous learning networks focuses on the function of messages as a whole and not the subject line. As a result it has been necessary to rely on previous research about email and other types of computer-mediated communication. As well as covering the potential functions and effects of the subject line, this section deals with a number of other factors affecting what messages students choose to open in a discussion forum. Students studying full time are often enrolled in a number of different courses at the same time and have a limited amount of time to spend on course activities. In addition, students studying part time usually have other commitments outside their studies and must therefore manage their time to their best advantage. In a study by Brown and Green (2009: 61) on the time spent reading and responding to discussion threads, it was estimated that graduate students spent between 50 and 76 min on the forum each week, and that students could compose a thread-initiating message and respond to two others in less than 2 h. The fact that there is limited time as well as a large number of participants and a large volume of messages posted in a forum means that it is not possible for participants to respond to every message. Therefore, students must make strategic choices about which messages to read and respond to.

In order for a text to be comprehensible it must be coherent (Schober & Brennan, 2003: 95). Coherence is somewhat difficult to define (Schiffrin, 1987: 21–29) and what constitutes coherent face-to-face communication may not be what constitutes coherent online communication, as the study by Schallert et al. (1996) implies. The term coherence is closely related to cohesion. According to Halliday and Hasan (1976) it has to do with the semantic relations within a text and can be expressed in different ways such as reference, substitution and lexical cohesion. It is also thought that cohesive devices contribute to coherence, although a lack of them may not make a text incoherent (Hellspong & Ledin, 1997; McNamara, Crossley, & McCarthy, 2010). Coherence might be said to refer to more than just the text itself and includes the context within which communication takes place (Collins & Hollo, 2000: 162).

The communication that takes place in a computer-mediated text-based discussion forum is normally multiparty and the participants do not share the same physical or temporal space. As a result of the communication being computer-mediated, there are potential obstacles such as lack of simultaneous feedback and disrupted adjacency (Herring, 1999). On the other hand, there are potential advantages of text-based computer-mediated communication such as conversational persistence (Herring, 1999). This means that the preceding exchanges remain visible and accessible. In written dialogue, there is time to edit and reflect on the contribution before it becomes known to the others (Tanskanen, 2006: 130). Participants appear to regard the persistence of text, the asynchronicity and the opportunity to edit as positive aspects (Hammond, 2000: 255). There are also communicative routines that carry meaning and make the interaction efficient (Condon & Čech, 2010). Even though turns may not always be adjacent to one another they can be connected in other ways. How a message connects to previously posted messages is “a central aspect of coherence in computer-mediated communication” (Severinson Eklundh & Rodriguez, 2004: 1). Linking a response textually to a previously posted message can be done in a number of ways, including reference to the author of the message, to previous messages, repetition, quoting, paraphrasing as well as cutting and pasting (Herring, 1999; Severinson Eklundh & Rodriguez, 2004; Tanskanen, 2006).

Lapadat (2007) uses a model for classification of discourse devices used to promote coherence. This model is based on the material that was analysed for a study where the different devices were identified and grouped into three types according to their deictic function: backward reference (reference to previous messages), forward structuring (reference to expected messages) and endogenous devices where the reference is to other contexts (Lapadat, 2007: 73). In a study of coherence in the classroom, Schallert et al. (1996) noted that in written conversation “it was rare for a comment to be posted without some sort of explicit marker showing how it was related to at least one comment that had been made previously” (Schallert et al., 1996: 480). In a comparison done by Paulus and Phipps (2008), they found, among other things, that naming and greeting were used somewhat more frequently in asynchronous than in synchronous discussions (Paulus & Phipps, 2008: 471). The two strategies of greeting and addressing other participants were often used in combination and could be found in the subject line or the message content. Addressing another person directly can have two functions. One of them is to single out the addressee among others and the function is to maintain social relationships (Biber, Johansson, Leech, Conrad, & Finegan, 1999: 1112). If a response is introduced with a greeting to a particular participant then it may help connect the response to that particular participant's message rather than the message in closest proximity (both spatially and temporally).

In order to place a message in a relevant context in the discussion it can be linked not only using the linking functions provided by the LMS, but also by textual linking. There are many textual links in messages and Herring (1996: 91) suggests that this is because the subject line alone is not enough to create coherence. There are some studies on the effect of the subject line in email correspondence, mostly from a marketing perspective, to find out how to increase the response rate. In one study on the effect of the subject line content of emails, Porter and Whitcomb (2005: 385) concluded that the subject line could not predict response rate. They attempted to find out how many potential students to a certain university would respond to surveys that were sent out via email. The message subject lines were categorised into three groups: request for help, name of school sending the message and blank subject line. Amongst the students who had already applied to go to the university in question, they found that the different subject lines had no effect on response rate. For the students who had not actually applied to go to that particular university, they found that the emails that got the most frequent responses among the three different categories were the emails with a blank subject line. This may be an indication that curiosity is a factor in predicting whether a message is read or not (Porter & Whitcomb, 2005: 384).

In a content analysis study on student interaction in computer conferencing, it was suggested that “efficiency and overall learning” would increase if students were instructed in the effective use of the subject line (Hara, Bonk, & Angeli, 2000: 147). By effective the authors mean that the subject line content should be both accurate and informative. They also conclude that providing the opportunity to label messages according to their discourse function might result in “cognitively deeper discussions” (Hara et al., 2000: 148). This view is not supported by another study of student messages where students were required to classify and label their messages in an attempt to improve efficacy. Firstly, it was difficult to label messages because most messages are polysemic. Secondly, the students that were required to label their messages were less likely to challenge the arguments of others and to respond to challenges, suggesting that the constraints imposed by the obligatory labelling may have inhibited learning and decreased the level of critical argumentation (Jeong & Joung, 2007: 441).

How accurately the subject line reflects the message content is a factor, which can potentially affect the strategic choices students make about which messages to access, read and respond to. The subject line can be a quick and effective way of reflecting stance in a discussion. In a study of conflict in computer-mediated communication (CMC), Graham (2007) gives examples of how participants use the subject line to reflect their position in a discussion, suggesting that this is an indication of the importance of this strategy for putting messages in a context. Graham also claims that participants who use this strategy appear as competent users of the medium, while participants who do not use a subject line that accurately reflects message content are seen as impolite (Graham, 2007: 748–750). An appealing subject line, which reflects interesting content, might also attract readers. Certain discussion topics are more popular than others in that they generate more interaction (Himelboim, 2008: 172). Zhao and McDougall (2008) indentified what they term “interesting topics” which affected participation. They also found that students tended to build on threads where many others were already posting messages (Zhao & McDougall, 2008: 69).

In order to answer the third research question concerning the effect of the subject line content on whether a message is opened, a number of other factors, which may also affect choices need to be mentioned. Researchers in CMC often use the term affordances to refer to aspects of the technology that may affect communication. This term is traced back to Gibson (1979) and is normally used to refer to the opportunities or capabilities of the technology, and what they allow the users to do or not to do. This does not mean that the technology decides how we communicate. Hutchby (2001) suggests that there are features inherent to the technology, which do not necessarily determine the way the technology will be used, but these features mean that the ways we can use the technology are somewhat limited. Examples of such features might be screen size or keyboard language and how discussion threads are organised. When studying CMC it is therefore necessary to discuss the technological affordances and the extent to which they may have an effect on the communication (Hutchby, 2001: 29). There are also a number of personal preferences that may influence interaction and these are discussed in Section 1.2.3.3.

The technical affordances of the LMS affect in what order the messages are displayed on the screen and how discussion threads are arranged. Messages posted earlier in the discussion are likely to be opened more times than the ones posted later, and messages at the end of a thread are also less likely to be read than those closer to the beginning of the thread (Thomas, 2002: 360–361). There is sometimes a tendency for students to post late in the forum, that is, just before the forum is closed. These messages are not likely to get a response from either the instructor or the other students (Gerber, Scott, Clements, & Sarama, 2005: 30; Thomas, 2002: 361).

One factor affecting whether a message is accessed or not may be who authored the message and how they are regarded by other participants. Certain participants, such as the instructor, may enjoy a higher status with regard to opinions and feedback and therefore it might be expected that their messages are more likely to be read than those of student peers. In one study on instructor influence in discussion forums, it was noted that there was no significant difference in the response patterns between the instructor and students with regard to which messages received a response (Gerber et al., 2005: 30). In another study of an open discussion forum containing mass messages it was found among other things that in order to cope with the volume of messages participants tended to choose a few popular participants to read and respond to (Himelboim, 2008: 163). Students have been found to use different strategies when choosing which messages to read in forums. Such strategies can consist of frequently logging in to avoid a backlog of unread messages, skimming messages, focussing on a single thread or selectively choosing or avoiding messages written by certain people (Peters & Hewitt, 2010: 954). Peters and Hewitt (2010: 956) also found that some participants read nearly all messages indiscriminately.

Because of the dialogic nature of an online discussion there is a possibility of some aspects of conversation analysis being relevant for the discussion of findings. An example of this is the effect of prior turn constraints (Sacks, Schegloff, & Jefferson, 1974). Even though participants in a dialogue are relatively free to say what they want in conversation, there are some constraints and these are determined by what the speaker before has chosen (Sacks et al., 1974: 711). With regard to factors such as turn-size and turn-order the conversation is seen as being “party administered” and “interactionally managed” (Sacks et al., 1974: 726). Communication is social action and is framed by the situation and type of community activity (Linell, 1998: 5). Linell (1998: 86) also states that the language we use when communicating “is deeply social and interactional in nature” and Sacks et al. (1974: 727) refer to the principle of recipient design when discussing the ways in which participants in conversation talk in a way that displays “an orientation and sensitivity to” co-participants. This orientation can manifest itself in topic selection, word selection, ordering sequences and options for starting and finishing a conversation. Branigan, Pickering, Pearson, and McLean (2010: 2358) suggest that choosing the same words or structures, as others may be a way to aid coherence. In one study of email communication among a group of students, it was revealed that group norms define communication patterns within groups and that variation is greater between groups than within them (Postmes, Spears, & Lea, 2000). Among other things Postmes et al. (2000: 351) found that there were significant differences between groups on the length of the message and references to the self.

Section snippets

Material and methods

Section 2 provides an overview of the material and methods used in the present study including the different groups, the number of participants and the number of messages posted in each forum. Factors such as the participants' commitments outside their coursework may also have had an effect on the extent of activity in the forum, but these are not within the scope of the present study. The present study is part of a larger study of student interaction in discussion forums where participants

Results

There were 119 thread-initiating messages posted at Level 1. A further twelve messages posted at Level 1 were incorrectly threaded responses to other thread-initiating messages. Therefore these twelve messages were analysed together with the responses posted at Level 2, making 194 response messages at Level 2. At Level 3, 73 student responses were posted, and 49 messages were posted at deeper levels.5

Limitations

The present study was an inductive one which attempted to identify types of subject lines as well as their function and also if there was a relationship between choice of subject line and whether other participants accessed and read the message or not. It has not been possible to draw any conclusions about the relationship between subject line type and the number who access a message. Primarily, this is due to the number of other factors that may also have affected the reading rate. The

Conclusion

The main aim of the present study was to investigate in what ways the subject line content contributes to overall coherence in an asynchronous online discussion in a learning environment and what other functions it might perform. The present study shows that subject line plays multiple roles, the most important being as an aid to coherence. The subject line referent can be a given title, the number of the course task or an idea expressed in the message. Thread-initiating messages primarily have

References (47)

  • D. Biber et al.

    Longman grammar of spoken and written English

    (1999)
  • A. Brown et al.

    Time students spend reading threaded discussions in online graduate courses requiring asynchronous participation

    International Review of Research in Open and Distance Learning

    (2009)
  • P. Collins et al.

    English grammar: An introduction

    (2000)
  • S. Condon et al.

    Discourse management in three modalities

    Language@Internet

    (2010)
  • D.R. Garrison et al.

    Critical inquiry in a text-based environment: computer conferencing in higher education

    The Internet and Higher Education

    (2000)
  • S. Gerber et al.

    Instructor influence on reasoned argument in discussion boards

    Educational Technology Research and Development

    (2005)
  • J.J. Gibson

    The ecological approach to perception

    (1979)
  • P.K. Gilbert et al.

    How to structure online discussions for meaningful discourse: a case study

    British Journal of Educational Technology

    (2005)
  • C.N. Gunawardena et al.

    Social presence as a predictor of satisfaction within a computer-mediated conferencing environment

    American Journal of Distance Education

    (1997)
  • M.A.K. Halliday et al.

    Cohesion in English

    (1976)
  • N. Hara et al.

    Content analysis of online discussion in an applied educational psychology course

    Instructional Science

    (2000)
  • C. Haythornthwaite et al.

    Community development among distance learners: temporal and technological dimensions

    Journal of Computer-Mediated Communication

    (2000)
  • L. Hellspong et al.

    Vägar genom texten

    (1997)
  • View full text