assessing participation in online discussion forums: a proposal for multidimensional analysis

13
- 137 - ISSN: 1133-8482 Píxel-Bit. Revista de Medios y Educación ASSESSING PARTICIPATION IN ONLINE DISCUSSION FORUMS: A PROPOSAL FOR MULTIDIMENSIONAL ANALYSIS EVALUACIÓN DE LA PARTICIPACIÓN EN FOROS DE DISCUSIÓN ONLINE. UNA PROPUESTA DE MULTIANÁLISIS Dra. Rosabel Roig Vila [email protected] Saulius E. Rosales Statkus [email protected] Facultad de Educación. Universidad de Alicante. Departamento de Didáctica General y Didácticas Específicas Campus de San Vicente del Raspeig. Ap. 99 03080, Alicante (Spain) El objetivo del presente trabajo es el de evaluar los foros de discusión de un curso de Especialista Universitario online sobre Tecnologías de la Información y la Comunicación de la Universidad de Alicante basándonos en la propuesta de análisis de Kay (2004). En este análisis es fundamental graficar la actividad de los foros para mostrar, una representación visual que facilite su estudio. La investigación ha permitido obtener resultados definitorios sobre la actividad del foro y ha aportado una propuesta de multianálisis en el ámbito de estudios sobre la evaluación de la participación en los ámbitos de comunicación definidos por los foros de discusión online. Palabras clave: evaluación, aprendizaje en línea, comunicación interactiva, aprendizaje activo This paper aims to assess the discussion forums being used in the programme Especialista Universitario online sobre Tecnologías de la Información y la Comunicación, taught at the University of Alicante, on the basis of the analysis model suggested by Kay (2004). In such a model, it is essential to represent graphically the forum activity so that the visual representation may improve analysis. This research has allowed reaching results which define the forum activity and has contributed with a proposal for multi analysis in the area of assessing participation in communication within online discussion forums. Keywords: Education, E-learning, Interactive communication, Active learning Nº 40 Enero 2012 - pp.137-149 ©2011 Píxel-Bit. 1. Introduction. Discussion and debate forums constitute an excellent platform for learning through moderated interaction between equals (peer to peer) and are based on the idea of the student being the center of the learning environment and oriented towards the development of critical thinking skills (Sahu, 2008). Peer to peer interactions provide a wide learning spectrum in which, besides the moderator’s orientation, students are exposed to their colleagues’ ideas and perspectives. The strength of these forums is based on

Upload: dcadeiua

Post on 17-Sep-2015

5 views

Category:

Documents


0 download

DESCRIPTION

Artículo de la revista Píxel-Bit

TRANSCRIPT

  • - 137 -

    ISSN: 1133-8482 Pxel-Bit. Revista de Medios y Educacin

    ASSESSING PARTICIPATION IN ONLINE DISCUSSION FORUMS: APROPOSAL FOR MULTIDIMENSIONAL ANALYSIS

    EVALUACIN DE LA PARTICIPACIN EN FOROS DE DISCUSINONLINE. UNA PROPUESTA DE MULTIANLISIS

    Dra. Rosabel Roig [email protected]

    Saulius E. Rosales [email protected]

    Facultad de Educacin. Universidad de Alicante.Departamento de Didctica General y Didcticas Especficas

    Campus de San Vicente del Raspeig. Ap. 99 03080, Alicante (Spain)

    El objetivo del presente trabajo es el de evaluar los foros de discusin de un curso de EspecialistaUniversitario online sobre Tecnologas de la Informacin y la Comunicacin de la Universidad deAlicante basndonos en la propuesta de anlisis de Kay (2004). En este anlisis es fundamental graficarla actividad de los foros para mostrar, una representacin visual que facilite su estudio. La investigacinha permitido obtener resultados definitorios sobre la actividad del foro y ha aportado una propuesta demultianlisis en el mbito de estudios sobre la evaluacin de la participacin en los mbitos de comunicacindefinidos por los foros de discusin online.Palabras clave: evaluacin, aprendizaje en lnea, comunicacin interactiva, aprendizaje activo

    This paper aims to assess the discussion forums being used in the programme Especialista Universitarioonline sobre Tecnologas de la Informacin y la Comunicacin, taught at the University of Alicante, on thebasis of the analysis model suggested by Kay (2004). In such a model, it is essential to represent graphicallythe forum activity so that the visual representation may improve analysis. This research has allowedreaching results which define the forum activity and has contributed with a proposal for multi analysis inthe area of assessing participation in communication within online discussion forums.Keywords: Education, E-learning, Interactive communication, Active learning

    N 40 Enero 2012 - pp.137-149

    2011 Pxel-Bit.

    1. Introduction.

    Discussion and debate forums constitutean excellent platform for learning throughmoderated interaction between equals (peerto peer) and are based on the idea of thestudent being the center of the learning

    environment and oriented towards thedevelopment of critical thinking skills (Sahu,2008). Peer to peer interactions provide a widelearning spectrum in which, besides themoderators orientation, students are exposedto their colleagues ideas and perspectives.The strength of these forums is based on

  • - 138 -

    Roig, R. y Rosales, S.E. Pxel-Bit. Revista de Medios y Educacin

    getting the student involved in learningactivities, encouraging interaction betweenequals, allowing the student to play the majorrole in the (learning) process and the teacherthat of the mediator-moderator of the activity(Sahu, 2008; Silva, 2004).

    Thanks to this process, students internalizethe orientations and guidelines of their morequalified colleagues when they write in acollaborative way (Garca & Perera, 2007). Eventhough student participation in a forum canbe assessed in different ways, the authorsagree that a more active participation doesnot necessarily imply a more insightful andcollaborative learning, or a better quality ofthe interventions (Gros & Silva, 2006; Ornelas,2007).

    There are some disagreements regardingthe use of discussion forums as a learningtool and their influence on achieving asignificant learning. Some scholars considerthem a ground-breaking tool and some othersthink their effective and significant use isminimal (Kay, 2006a). The main reason behindthis divergence could be the lack of a solidand comprehensive tool to assess forums,which should be consistent, integral andbased on theoretical concepts (Kay, 2006b).

    Based on an extensive revision of previousresearch, Kay uses the following variables toassess traditional discussion forums: thread,position of the message within the thread,author (learner vs. educator), clarity of thesubject field, time of publication, lapse of timesince the previous message, number of timesthat a message has been read, number ofwords, main purpose, quality of the message,level of complexity of the subject, level ofknowledge and level of usage and processingof external resources. All these variables havebeen put in conjunction in a table, and foreach of them we have specified the possible

    assessment values. These variables haveproven to be effective in the assessment ofthe 12 key areas of discussion forum usage.Kay claims this type of metric is essential toimprove our understanding of onlinediscussion boards.

    2. Methodology.

    2.1. Research design.

    This study has been developed withregistered information from an activity carriedout in the discussion forum of theSpecialization Program on Applied ICT inEducation taught at the University of Alicanteand developed on the universitys Eduonlineplatform. Because of its characteristics, theresearch is descriptive and does not intend togeneralize results (Arias, 1999). The main aimis to use a predefined assessment tool in orderto describe the activity in virtual discussionforums.

    Using Kays metric (2004 and 2006) toanalyze virtual discussion forums, we havecarried out the assessment in two parts: onededicated to describe the forum using Kaysmetric (and to understand the metric better,adapting it if necessary to a specific universityenvironment) and to suggest a way ofrepresenting the forums activity graphically.We have decided to use this metric and toperform a manual assessment. However, wehavent ruled out the possibility of developinga similar research study in the futureperforming a qualitative assessment (withtools such as AQUAD, NUD*IST, etc.)

    2.2. Participants.

    In order to carry out the investigation wehave taken into account the total number of

  • - 139 -

    ISSN: 1133-8482 Pxel-Bit. Revista de Medios y Educacin

    students (N=17) and teachers (N=6) in theaforementioned Specialization Course. Theseparticipants have published a total number of720 messages, out of which a representativesample has been selected.

    The student group was formed by graduatestudents of different disciplines andoccupations, from both a scientific and ahumanities background, ranging fromcomputer programmers or IT engineers tosecondary school music teachers, socialsciences teachers, or English and primaryschool teachers. Once the size of our samplewas clear, we selected the items (messages)which were meant to be our object of analysis.The sampling has therefore been made in anon-probabilistic and unintentional way,following some criteria based on the revisionof related bibliography.

    The course is divided into topics and foreach topic several discussion threads weresuggested and developed. Based on thisstructure, we have selected arbitrarily onethread for each of the topics. The itemsselected are the ones which show a significantsocial learning based to the level of activity,student and teacher participation and numberof active participants in the discussion (Kay,2004).

    2.3. Data collection.

    The data has been obtained following twodifferent procedures 1) logging into the onlinecourse and then entering the online discussionforum; 2) accessing Eduonlines databasetables for this specific course. These tablesinclude data about the topics, threads, numberof messages, message sequencing within thethread, message author and message date,among others. Due to the point in time in whichthis study has been planned and developed,

    we have not conducted any surveys as theanalysis has been performed after the closureof the course.

    The assessment has been carried out withtwo perspectives: a quantitative and aqualitative one. The data received about forumparticipation has been treated with MicrosoftExcel in conjunction with SPSS Statistics 17.0and Statistica 8.0 for the graphs.

    2.4. Forum activity graph: a proposal foranalysis.

    When trying to represent the differentinteractions established in a discussion forum,there is an element which must be taken intoaccount which acts as a unifier of theparticipant interaction. This element is thediscussion thread. Therefore, taking as astarting point the article by Rallo and Gisbert(2008) and taking into account the importanceof the thread in the current study, we suggestthe following guidelines to construct thegraphs which will to show visually the boardsactivity: (1) There are two types of nodes:Type 1: participant (user); type 2: thread; (2)All the nodes represented in the graph haveat least one edge; (3) The type 1 nodes (user)are represented with an ellipsis or a roundcorner square; (4) The type 2 nodes (thread)are represented with a straight corner square;(5) The type 1 nodes (user) are labeled withthe name of the user which might be either astudent or a teacher. The type 2 nodes arelabeled with the thread name, taken from thesubject field, or the thread number, taken fromits identification number in the originaldatabase; (6) Type 1 nodes and type 2 nodesare presented in a different colour so that theycan be easily distinguished (optional). In thisstudy the student participants are alsodifferent in colour from the teachers or

  • - 140 -

    Roig, R. y Rosales, S.E. Pxel-Bit. Revista de Medios y Educacin

    tutors; (7) If there is a connection betweentwo different type 1 nodes this will beindicated through a type 2 node, which is tosay the users are interacting betweenthemselves through the threads; (8) The edgesare not directed and therefore the graph is notdirected; (9) The size of the nodes (of bothtypes) will increase with the number of edgeswhich come to (or go out of) them. The nodeswith more activity will therefore be bigger andwill stand out visually; (10) User participationis measured by the number of edges in thenode representing it. (11) The interestgenerated by the thread can be measured byits number of edges. The more edges thereare in the node, the more interesting we canconsider the thread to be, as it has stimulatedmore participation. There are different waysto indicate who started a thread: (1) Adding inthe corresponding node the numbers of thethreads started by that user; (2) Adding thetag 1 in the edge that connects the user tothe thread. (3) The edges can be numberedwith the participation order of the users in aparticular thread. By doing this we will be ableto track whenever a user took part in thediscussion. Multiple posts by a user in thesame thread can be represented as follows:(1) Each participation generates a new edgetowards the thread; (2) An only edge is shownand we increase its thickness with the numberof interventions; (3) A type 2 node (thread)with one edge represents a thread which didnot get any replies; (4) The most heavynodes are placed towards the centre of thegraph (optional).

    2.5. Forum content: the thread selection.

    In order to analyze the forum by surveyingthe content of the posted messages, we havetaken as a starting point the graphic

    representations previously shown, from whichwe have selected the threads which seem tobe more interesting according to the numberof active participants and the number ofmessages published in it.

    These selection criteria allow us to observetwo different dimensions: when the criterionis the number of edges connected to a thread,that is to say, the amount of messagespublished, we can assess whether thisdiscussion has been active or not. The biggerthe number of messages, the more intense theactivity is, and we can therefore say that thethread reflects the development of aninteresting discussion. On the other hand, thenumber of participants which have taken partin a discussion comes to reflect a bigger (orsmaller) social reach of a particular discussion.If the number of participants is big, the topicis considered of interest for a larger numberof users which have wanted to take part in itand make a contribution. In any case, bothdimensions can co-exist and, in fact, a threadwith a high number of participants publishinga substantial amount of messages should bereflecting the personal and social growth ofthe group and the personal growth of itsmembers, and should be an indication of ameaningful and notable discussion in itssocio-educational dimension.

    Another key element which has been takeninto account is the subject field. Sometimesa thread with a more intense activity (highernumber of edges) has only made a smallcontribution to the course. An example of thisis the thread number 3991: the topic aboutmailing lists is the most active one (accordingto its number of edges: 8 messages) and theone with the highest number of participantsinvolved (n=5), but the content of thediscussion only serves to find an answer to aspecific doubt about how to use a computer

  • - 141 -

    ISSN: 1133-8482 Pxel-Bit. Revista de Medios y Educacin

    application. Another negative aspect of thisthread is the lack of a close up. One positiveaspect about it, though, is related to its socialdimension, as several students offered theirhelp in order to solve another studentsproblem.

    3. Results.

    A total of 720 messages have beenpublished (see Graph 1), distributed in 237threads on about 15 topics (M=15, SD= 8.72,Min=5, Max=37 threads per topic).

    The published messages (720 in total) aredistributed between 15 topics and have anarithmetic average of 48 messages per topic(SD= 17.94, Min=14, Max=82 messages pertopic). The average thread length (measuredby the number of messages in it) is 3 messages.The maximum thread length is 24 messages(on the topic of digital blackboards).

    The average word count per message is91.64 (SD=101.52, Min=1 word andMax=989 words).

    The average response time to themessages is 3.39 days (SD = 10.71; Min = 0;Max = 73). A total of 96 messages have beenanswered on the same day (13.33%) and a total

    of 85 messages have been replied to thefollowing day (11.80%). In some cases therehas been an extreme response time such as66, 79, 54 or even 118 days.

    The number of threads which have fiveor more messages is 34 (14.35%). In 48 threads(20.25%) the students and teachers havetaken part more than once in the discussion,with a maximum of 7 comments made by thesame person.

    The messages published by the teachersconstitute a total of 29.63% of the messagesand represent almost a third of theparticipation. The introductory topics to thecourse have a teacher participation rate of47.22% and a topic about Webquest inquirieshas a 42.86% (see Graph 2). The peculiarity ofthe presentation topic is that the teacherreplies to each of threads started by studentswith a welcome message.

    3.1. The forum seen through graphs.

    Following the guidelines explained in theprevious section we can represent graphicallythe discussion forums activity, taking intoaccount the more relevant elements: users,threads and the interrelations between them.

    Graph 1. Number of threads per topic

  • - 142 -

    Roig, R. y Rosales, S.E. Pxel-Bit. Revista de Medios y Educacin

    The graphical representation of the threadsin this particular forum has been createdmanually, but supported by a computerapplication. Therefore, we have generated thegraphs in two stages: in the first stage wehave revised every single course thread andhave generated the first outlines manually,following the guidelines stated above. Then,

    with the help of the tool Cmap Tools developedby the Institute for Human and MachineCognition in Florida, we have digitalised themanually-created outlines. Cmap Tools is anapplication which allowed us to work withconcept maps effectively.

    These are some examples of the threadgraphs:

    Graph 2. Students/Teachers forum posts

    Figure 1. Forum participation - Topic 0

    Topic 0: Course welcome.

  • - 143 -

    ISSN: 1133-8482 Pxel-Bit. Revista de Medios y Educacin

    This activity involved posting a briefmessage in the forum about the studentsforeknowledge and experience, etc. As we cansee in Figure 1 this was basically a one to oneinteraction between each participant and thecourse tutor, whose answer was a welcomemessage to the course.

    In this topic there is a very discreet teacherparticipation (see Figure 2), which can beinterpreted as a positive strategy to let thestudents be the protagonists of the discussionand the learning. Moreover, a more interestingdiscussion structure can be noted due to thehigh number of students involved in thethreads with a more active participation. Wecan also appreciate a possible trend whoeverstarts the thread is usually the most activeparticipant in the thread. It could be said thatthe users who start the threads feel in a wayresponsible for their evolution or that theymight have to stand up for the ideas originallystated. After observing this diagram we couldsay that in this topic the threads play a fairly

    bigger role than the participants.In the topic about teacher roles, we can

    again appreciate an important participation ofthe moderators, and an interest in it from apart of the student group. Moreover, as wecan see in Figure 3, three of the users involvedin the discussion made up to 4 contributionsto it. The rest of the threads are lessoutstanding. It is interesting to see in the graphthe interaction between all the students,regardless of the threads that got no reply orwere not very active. Practically all thestudents registered in the course were activein this topic, which makes it an illustrativeexample of social learning.

    We can generally observe that the forumhas been used a direct asynchronouscommunication tool between two interlocutorswhich also allowed the rest of the students tofollow the discussion. When seen in thegraphs, we can appreciate that there are veryfew threads with a discussion involving allthe students actively.

    Topic 2: ICT and education in the society of information.

    Figure 2. Forum participation - Topic 2

  • - 144 -

    Roig, R. y Rosales, S.E. Pxel-Bit. Revista de Medios y Educacin

    We must also emphasize that in most topicsit was one of the teachers who participatedmore actively in the forum, taking part inalmost every thread and leading, maybeexcessively, the current discussion.

    3.2. General analysis of the selectedthreads.

    After carrying out an assessment of eachof the selected threads, we have displayedthe results in a table and have performed abrief general analysis.

    Clarity of the Subject field: in most cases(n=87; N=140; 62%) the threads subject fieldis not clear enough and is usually made up oftwo or less words. It seems to be a commonpattern to use two words in the subject fieldwhich do not indicate clearly the content ofthe message: 13% (n=18) are somewhat clear,19% (n=27) are quite clear and 6% (n=8) arevery clear.

    Main purpose of the message: (see Graph

    3) following the pattern of the field questioneasily answered by using other sources, 76%of the messages (n=105; N=140) are answersto questions asked at the start of the threador to activities suggested as course material.7% (n=10) of them are open responsequestions, presumably seeking aninvolvement in the activity of all the membersin the group even though, regardless of thequestions complexity, as we have seen in theprevious section: 7 messages (5%) involveda student query about the discussion; 16messages (12%) are independent andunrelated to the thread discussion and justone message (1%) of the ones underassessment has been a comment unrelated tothe discussion and with no educational value.To sum up, practically all the messages wererelated to the course, regardless of theircontribution (which we will discuss later).

    New information added: almost half of themessages (n=65; N=140; 46%) do not addany new information to the discussion.

    Topic 3: Functions, roles and competences in the current technological context.

    Figure 3. Forum participation - Topic 3

  • - 145 -

    ISSN: 1133-8482 Pxel-Bit. Revista de Medios y Educacin

    Sometimes the messages are published inorder to support or confirm somethingpreviously stated. Other times, once thediscussion threads have become lengthy, themessages have become repetitive and haveacquired a social dimension because of theuse of greetings and anecdotes in them. Thesemessages are usually related to the topic butthey dont always add any new knowledge:15 messages (11%) add some knowledge inan indirect way and 60 messages (43%), quitea considerable percentage, provide an answerto the questions or activities suggested.

    Quality of the message: in 19 (14%) caseswe havent found in the table the conditionsrequired for the messages to be assessedwithin the scale and there is no incorrectmessage: 20 messages (14%) provide someinformation which is not related to thediscussion topic or to the course; almost halfof the messages (n=69; 49%) are acceptablebecause they reply in some way to thequestion or to one of the suggested activities;24 messages (17%) are valid because theyrespond to a part or to the whole threadquestion; only 8 messages (6%) are excellent

    as they provide an answer in a clear andrelevant manner and also add significantdetails. These excellent messages have beenpublished by the teachers. On severaloccasions the questions and activities wereopen and they encouraged the students togive their opinion or reflect on their ownexperience. In such situations, the answerscould be taken as correct, but depending onhow pertinent they were and on theinformation they provided they were markedas either good or acceptable. Very few threadshad specific questions, and that is why theassessment of this line should be consideredto be very subjective and dependent on theexperience and knowledge of the assessor.

    Type of knowledge: due to the nature ofmany of the questions and activities, there isa high percentage (41%) of messages (n=57;N=140) which contribute with facts(anecdotes and comments with littleelaboration, among others); however, 27% ofthe messages (n=38), are more elaborated anddraw a connection between two or more facts;9% (n=13) of the messages provideinformation about which procedure to follow

    Graph 3. Main purpose of message (all of them)

  • - 146 -

    Roig, R. y Rosales, S.E. Pxel-Bit. Revista de Medios y Educacin

    in order to carry out a specific task. Thesemessages are generally replies to equallyspecific and punctual questions. Finally, a12% (n=17) of the messages have beencharacterized as being metacognitive, becausethe students are reflecting on how to find apossible solution to a problem. This last resultcan be seen as a very positive aspect whichsuggests evidence of the students ability toprovide solutions.

    External resources: in general terms linksto external sources are rare in the messages(n=97; N=140; 69%). In some cases (n=12;9%) the course information or a post writtenby a teacher is referred to; 16% (n=23) makereference to the information posted in anothermessage; 3% (n=4) mention a website.

    We can interpret this as a result of theteachers having provided enoughinformation to cover the topics, but it couldalso be a sign of the students either not beingfamiliar with making references to externalsources or not making the effort to do anyfurther research on the topic beyond theinformation provided.

    We must also take into account that somethreads were asking the students to narratetheir personal experiences with the use of ICTon their own work, while in other threads theyhad to comment on the material provided.

    4. Discussion.

    The metric suggested by Kay (2004) hasbeen very useful in order to systematize theanalysis of forum activity. Kays guide ispractical in nature and allowed us to structureour assessment, and to take into accountmany variables which are put into practice inthreaded discussion forums. However, we hadto make a prior revision and updating of theguide in order to define the variables that could

    be assessed in the forum. Despite having ruledout some of these variables, by using Kaysguide we had to add the value NA (Does NotApply) for some fields which did not apply tothe messages under study. Even though theguide allowed us to systematize the analysis,it should be noted that the assessment of thecontent was at the discretion of the researcher.This manual task involved reading each andevery thread message, assessing some of theircharacteristics and attributing different valuesto them according to Kays table. In somecases the message was in the borderline divingthe two categories, and the assessor had tolean towards one of the two options. There isa level of subjectivity in this task and itthoroughly depends on the assessorsinterpretation of the guide and messages.Performing this work manually implies asubstantial effort as the selected discussionmessages have to be read several times inorder to assess all the variables at play, and todetermine in which category they should beplaced in. At the beginning of the analysisthe assessment is even more complicated asthe assessor has not yet memorized all theindicators that allow evaluating the messageaccording to the metric. This situation getssolved as soon as the forum assessmentproceeds and it eventually evolves into amechanical dynamic.

    After carrying out this manual assessment,we consider of great importance the use of ameasuring tool in order to organise andexpedite the process. We also wonder howwould the analysis be carried out if we wereto emulate Ezeiza and Palacios (2009) and doit in a semiautomatic (or even automatic?) wayusing software tools such as NUD*IST orAQUAD.

    One of the aims of the current study is tosuggest and we say this with all humility- a

  • - 147 -

    ISSN: 1133-8482 Pxel-Bit. Revista de Medios y Educacin

    good way of showing graphically the activityof a thread so as to facilitate its visualinterpretation. In these graphs we can clearlysee the more active threads, which threadsoutstand from the rest because of their activity,which user has taken part in a discussion,which threads didnt get any replies, etc.

    All in all, the guide to turn data into graphshas been very useful in order to begin theforum analysis, but we recommend its revisionand improvement, as well as encouraging itsregular use in the analysis of virtual discussionboards. On this topic, we must also say thatthat even though Cmap Tools, the applicationused to generate the graphs, has been of greathelp, we consider it necessary to look into thecreation of new software designed specificallyfor this purpose. A starting point in the creationof such a tool could be the analysis of theinformation registered in the database andusing each register to create the structure ofthe graph automatically.

    Generally speaking, and regarding theforum analysed, we would like to reflect onsome interesting observations. First of all, wethink the students havent made the most ofthe learning possibilities which online forumscan provide, as we have seen in the theoreticalframe of this study. An average of threemessages by thread is a considerably lownumber of messages. According to Kay (2004)social learning happens after the 5th message,and this situation has only happened in 14%of the discussions. It seems a common practiceamong the students to start a new threadevery time they have a contribution to make,which implies that they have not beenfollowing the discussions closely. In generalit can be perceived that the forum has beenused as a direct asynchronouscommunication tool between twointerlocutors, where the conversation can be

    read by all the members of the group. Whenseen graphically, there are not manydiscussion threads in which every user getsactively involved.

    Sometimes there is a very active teacherparticipation in the discussion. We mustremember that the student should play theleading part in the learning process (Garca &Perera, 2007) and that the role of the tutorshould be to stimulate the group, organizingthe activities, motivating the students,creating and sustaining a pleasant discussionenvironment and facilitating the teaching-learning process (Silva, 2004).

    The quality of the messages published(according to Kays table) is overallacceptable (49%) and only 6% of them areexcellent. Based on Garca and Perera (2007)this is not rare, as encouraging the studentsto share their ideas to build their knowledgethrough debate entails a considerable degreeof difficulty.

    We can also appreciate after having madethe analysis with the graphs and the aid ofKays metric that, as noted by Gros and Silva(2006) and Ornelas (2007), discussions with amore active participation are not necessarilythe ones with a better quality. We can observein this study after using Kays metric thatthreads with an interesting participation didntshow any quality messages or arguments. Inaddition to this, many of the first messages inthe threads were interesting and focused onthe initial problem, while the rest were socialcommentaries and anecdotes (without takinginto account the threads which involvednarrating their personal experience with ICTin learning).

    In any case, there are some examples inwhich the students have made the most ofthe forums possibilities. In topic 2 (ICT andeducation in the information society), threads

  • - 148 -

    Roig, R. y Rosales, S.E. Pxel-Bit. Revista de Medios y Educacin

    3988 and 3994 are good examples of what sociallearning should be like, as it allows the studentto play the leading role and to internalise theorientations and guidelines from the most ablestudents when writing in a collaborative way(Garca & Perera, 2007). In these cases, theinitial topic questions asked by the teacherhave encouraged dialogue among studentsas well as social learning. Later on, the teacherhas made a couple of interventions to keepthe discussion active and to clarify someaspects so as to lead the conversation backto its original purpose. This could be read asthe teacher allowing the students to be incharge of the conversation, always keepingtrack of its development but stepping backand making contributions only whennecessary.

    In order to make the best use of onlinediscussion forums in the learning environment,it is advisable that the teachers and tutorsshould previously define the characteristicsand the use they will make of the forum(inquiries, discussion, bulletin board, etc.) andplan in detail the topic that will be discussedin the forum, the orientation the discussionshould be following and, most of all, wherethe discussion should lead to (conclusionsand thread close up). This way the teacherwill be able to guide and moderate thediscussion and make the minimal interventionsnecessary in order for the students to be theones who construct knowledge and be thekey of the learning process.

    On the other hand, it is recommended thatstudents should lose fear to participate andcover the topics in depth. As we havementioned before, according to Garca andPerera (2007) it is very difficult to makestudents go from sharing their ideas tobuilding knowledge, and therefore thestudents should be aware of their

    responsibility in the learning process andremember that the better the quality of theirinterventions in the discussions, the betterresults they will obtain.

    We wouldnt want to bring this study to anend without remarking that communication isessential in the teaching-learning onlineprocesses. Beyond technology, it is humancommunication and the changes in the socialinteraction patterns which are the key (Roig,2009). Taking this consideration as a startingpoint, discussion forums could then be validtools in order to establish the necessarycommunications channels needed to achievea significant learning.

    5. Research funding sources.

    The present study is framed within theResearch Project EDUTIC-ADEI (Ref.:Vigrob-039), the Research Group onInnovation in Education Technology (GITE)EDUTIC-ADEI-EDAFIS, and the NetworksProgramme of Investigation and UniversityTeaching of the ICE, all of them based in theUniversity of Alicante (UA); of the Projecte-Accesible (LIA for Deployment andInternationalization of the System, GeneralSub directorate for Public-Private PartnershipStrategies, Subprogram INNPACTO, SpanishMinistry of Science and Innovation, Ref. IPT-430000-2010-29 (2010-2013)), of the projectIVITRA (http://www.ivitra.ua.es) and theProject DIGICOTRACAM (ProgramPrometeo of Research Groups of Excellent,Ref.: Prometeo-2009-042, co-financed by theERDF of the EU and the Spanish Ministry ofScience and Innovation FFI2009-13065.)

  • - 149 -

    ISSN: 1133-8482 Pxel-Bit. Revista de Medios y Educacin

    6. Bibliographical references.

    Arias, F. (1999). El proyecto de investigacin.Gua para su elaboracin. Caracas: Episteme.Ezeiza, A., & Palacios, S. (2009). Evaluacinde la competencia comunicativa y social enforos virtuales. RELIEVE. Revista electrnicade investigacin y evaluacin educativa, 15,1-15. Retrieved from http://www.uv.es/RELIEVE/v15n2/RELIEVEv15n2_2.htmGarca, C. & Perera, V. (2007). Comunicacin yaprendizaje electrnico: la interaccindidctica en los nuevos espacios virtuales deaprendizaje. Revista de Educacin, 343.Retrieved from http://www.revistaeducacion.mec.es/re343/re343_17.pdfGros, B. & Silva, J. (2006). El problema delanlisis de las discusiones asincrnicas en elaprendizaje colaborativo mediado. RED.Revista de Educacin a Distancia, 16.Retrieved from http://revistas.um.es/red/article/view/24251/23591Kay, R. (2004). Developing a Metric forEvaluating Discussion Boards. En J.Nall & R.Robson. (Coord.). Proceedings of WorldConference on E-Learning in Corporate,Government, Healthcare, and HigherEducation 2004. (pp. 1946-1953). Chesapeake,VA: AACE.______ (2006a). Developing a comprehensivemetric for assessing discussion boardeffectiveness. British Journal of EducationalTechnology, 37, 761783._____ (2006b). Using asynchronous onlinediscussion to learn introductory programming:An exploratory analysis. Canadian Journalof Learning and Technology, 32(1). Retrievedfrom http://www.cjlt.ca/index.php/cjlt/article/viewArticle/64Ornelas, D. (2007). El uso del Foro deDiscusin Virtual en la enseanza. Revista

    Iberoamericana de Educacin, 44, 1-5.Rallo, R. & Gisbert, M. (2008). Anlisis de unacomunidad on-line a partir de su lista dediscusin. El caso de Edutec-L. Edutec.Revista Electrnica de TecnologaEducativa, 25. Retrieved from http://edutec.rediris.es/Revelec2/Revelec25/Edutec25_analsis_comunidad_online. htmlRoig Vila, R. (2009). Redes sociales ycomunidades virtuales en la Web 2.0.Implicaciones en el mbito educativo. En RoigVila, R. (Dir.). Investigar desde un contextoeducativo innovador. (pp. 399-412). Alcoy:Marfil.Sahu, C. (2008). An evaluation of selectedpedagogical attributes of online discussionboards. Hello! Where are you in thelandscape of educational technology?Proceedings ascilite Melbourne 2008.Retrieved from http://www.ascilite.org.au/conferences/melbourne08/procs/sahu.pdfSilva, J. (2004). El rol moderador del tutor en laconferencia mediada por ordenador. Edutec.Revista Electrnica de TecnologaEducativa, 17. Retrieved from http://edutec.rediris.es/ Revelec2/revelec17/silva_16a.pdf

    Reception date: 2011-04-24Assessment date: 2011-06-20Acceptance date: 2011-06-29Publication date: 2012-01-01

    /ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile () /PDFXOutputCondition () /PDFXRegistryName (http://www.color.org) /PDFXTrapped /Unknown

    /Description >>> setdistillerparams> setpagedevice