Do Virtual Groups Recognize Situations in Which It Is Advantageous to Create Digital Concept Maps?

In the present experimental study, it was investigated whether virtual groups, depending on their situational circumstances, were able to decide in favor of the more suitable problem-solving procedure for their situation. The situational circumstances were to either have the possibility to create a “knowledge and information awareness” approach or not, that is, an approach that provides to the group members their collaborators’ knowledge and information by means of digital concept maps. The study compared 20 triads with spatially distributed group members that were able to create a “knowledge and information awareness” approach with 20 triads collaborating without this possibility. Results showed, as expected, that the triads mostly chose the more suitable problem-solving procedure for their situation and that deciding in favor of the more suitable procedure resulted in both less time needed for solving the problems and less perceived coordination effort. However, triads that were able to create a “knowledge and information awareness” approach often did not finish their approach and could therefore not benefit from the full potential of this approach. The results are discussed. 1 Theoretical Background 1.1 The Benefits of Knowing What the Collaborators Know Different fields of research show us the importance of knowing what the collaborators know in order to be able to communicate and collaborate effectively in group situations (Engelmann & Hesse, 2010). The research on audience design (e.g. Dehler-Zufferey, Bodemer, Buder & Hesse 2011) gives reason to believe that knowing what the collaborators know leads to changes of behavior such as writing longer texts about a topic when addressing novices (e.g., Dehler, Bodemer, & Buder, 2007). The knowledge imputing approach by Nickerson (1999) points out that knowing what the communication partner knows improves the communication by avoiding possible misunderstandings. The theory of transactive memory system (Wegner, 1986, 1995) states that group members need to know which member possesses knowledge about which topic in order as to access it through communication. An effective transactive memory system has shown to improve group performance (e.g. Liang, Moreland & Argote, 1995). This knowledge about the collaborators’ knowledge is, however, quite difficult to acquire: For example, groups need enough time to acquire it (e.g. Wegner, 1986) and many different mistakes may occur during the acquisition process (Nickerson, 1999). The acquisition of such knowledge is especially challenging for virtual groups, that is, in cases in which the spatially distributed members have to collaborate computer-supported; these groups have to struggle with reduced contextual information caused by the use of computers (cf. Kiesler, Siegel, & McGuire, 1984). The “knowledge and information awareness” (KIA) approach (e.g. Engelmann, Tergan, & Hesse, 2010; Engelmann & Hesse, 2010; Schreiber & Engelmann, 2010; Engelmann & Hesse, 2011) is a proven solution for these kinds of problems. 1.2 The KIA Approach and Coordination Theory According to Engelmann and colleagues (e.g., Engelmann & Hesse, 2010), KIA is defined as being informed with regard to the collaborators’ knowledge structures and underlying information. KIA is fostered by providing the group members with access to both their collaborators’ knowledge structures and underlying information both visualized via digital concept maps. In the study these concept maps consisted of task-relevant labeled concepts as well as relations between those concepts which, as a whole, embodied both the knowledge structures and the information of the collaborators. Elements with additional information were linked to the concepts and could be opened in a small desktop window by mouse-clicking. In several studies, this approach has been empirically proven to foster KIA acquisition of spatially distributed group members who collaborated computer-supported (e.g., Engelmann & Hesse, 2010; Schreiber & Hesse, 2010). Besides this, the KIA approach was also confirmed to improve computer-supported collaborative problem-solving: It enhanced group-performance in simulated virtual triads, meaning three persons sitting in the same room, able to speak with each other but not able to see each other because of partition walls (Engelmann et al., 2010). A follow-up study with real virtual groups, meaning groups with spatially distributed members, replicated these findings (Engelmann & Hesse, 2010). Evidence for this effect was also found with more complex tasks and different knowledge domain material (Schreiber & Engelmann, 2010). One reason for the effectiveness of this approach is that it facilitates coordination. According to Malone and Crowston (1994), coordination is the management of dependencies between activities. Especially in the field of collaborative problem-solving and knowledge management, the management of producer-consumer relationships is very important. Producer-consumer relationships, as Malone and Crowston (1994) state, often lead to special kinds of dependencies such as ‘prerequisite constraints’ or ‘transfer’. ‘Prerequisite constraints’ are activities that have to be finished before other activities can be started. For example, before person B is able to make a decision or solve a problem, person A has to provide the needed information. ‘Transfer’ takes place in between, when the producer communicates or ‘transfers’ information to the consumer. The coordination or management of these dependencies in a group problem-solving setting entails effort or costs, such as time costs for prerequisite constraints or the need to correct mistakes that appeared in the information transfer process. If the coordination costs are low, the coordination process is effectively. It is expected that the KIA approach improves coordination: In the process of group knowledge building as a requirement for group problem-solving, the KIA approach provides the opportunity to make parts of the individual knowledge of the group members available before the individual knowledge building process is completed. Parts of the individual knowledge as a product can be used by others before the entire product is completed. This should lead to a time advantage compared to groups whose members cannot see parts of others individual knowledge before it is completed. Moreover, having the possibility to see the knowledge of the collaborators makes it possible to use this knowledge without the need to ask for it, or in return, to communicate it. This should lead to both fewer mistakes and less time needed for completing collaborative tasks. As a result of such process costs reductions, the collaborative problem-solving of groups provided with the KIA approach should be more effective and efficient. 1.3 Bringing the KIA Approach into Practice Prior studies of Engelmann and colleagues (e.g. Engelmann & Hesse, 2010) were highly structured and the utilization of the KIA approach was predetermined because the group members worked with maps that were created by an expert (instead of by the group members themselves). These expert maps contained the complete content needed to solve the problems. Individual and collaborative work phases were separated from each other and offered enough time to accomplish each task. This design differs from real application fields in which group members need to externalize their individual knowledge by themselves. Moreover, groups need to decide in favor of a suitable, coordinated problem-solving procedure, depending on the situational circumstances (West, 1996). They also need to adapt their activities to the situational or environmental circumstances as a requirement for a coordinated procedure (Salas, Sims & Burke, 2005). Such environmental circumstances are, for example, technical characteristics and affordances. In the present study, the situational circumstances were to either have the possibility to create a KIA approach or not. This leads to our research question: Depending on whether the groups in an unstructured situation have the possibility to create a KIA approach or not, do they opt for a procedure that is suited to effective, computer-supported collaborative problem-solving? 2 Experimental Study The present study compared a control condition with an experimental condition. The groups in the experimental condition had the possibility to create a KIA approach, meaning that each group member could first visualize his/her own knowledge and information by means of a digital concept map which afterwards was provided to the other group members during a collaborative problem-solving phase. In contrast, the groups in the control condition did not have this possibility. These group members could create individual concept maps, but these maps would not be shared with the other group members during the collaborative problem-solving phase. We expected that groups, according to situational circumstances, would decide in favor of a suitable problem-solving procedure: The suitable procedure for the experimental groups would be that all group members, after having started to create their own concept map, continue with their individual maps before starting to collaborate, that is, to continue finishing the representation of their own knowledge and information. This is because the completed individual maps provided to the collaborators would function as a KIA approach that improves collaborative problem-solving. In contrast, the suitable procedure for the control groups would be that the group members start directly with collaboration because they cannot see the completed partner maps and, therefore, cannot create a KIA approach. Thus, we propose the following hypotheses: • Hypothesis 1: Groups, depending on situational circumstances, that is, depending on the possibility to create a KIA approach, choose the more suitable problem-solving procedure for their situation, which means that groups in the experimental condition decide more often to continue first with their individual maps while groups in the control condition decide more often to start directly with collaboration. • Hypothesis 2: Groups in the experimental condition that first continue with their individual maps and groups in the control condition that start directly with collaboration need less time for collaboration than those groups in the experimental condition that do not continue with their individual maps and those in the control condition that do not start directly with collaboration. • Hypothesis 3: Groups in the experimental condition that first continue with their individual maps and groups in the control condition that start directly with collaboration perceive less coordination effort than those groups in the experimental condition that do not continue with their individual maps and those in the control condition that do not start directly with collaboration. Prior studies have confirmed that a completed KIA approach, that is, having access to the concept maps of the collaborators, visualizing their complete knowledge and information, improves collaborative problemsolving. Therefore, we hypothesized the following: • Hypothesis 4: Groups in the experimental condition that first continue with their individual maps outperform groups in the experimental condition that directly start collaboration and groups in the control condition that either continue or do not continue with their individual maps.