Improving Virtual Team Collaboration Outcomes Through Collaboration Process Structuring

Over the last decade the defining factors and motivations behind how we work and how we learn have significantly and steadily moved toward a globalized network that encourages the sharing and creation of knowledge and information. The result is an increase in the number of universities offering courses and degree programs in an online environment and an increase in the number of corporations that employ and train knowledge workers in a virtual, distributed environment. According to a study completed in 2008 by The Sloan Consortium, the number of students taking at least one online course in the fall of 2007 was 3.94 million, which is an increase of 12.9% over the fall of 2006 (Allen & Seaman, 2008). This study, which is based on responses from more than 2,500 colleges and universities, also concludes that global economic factors are a primary influence on increased online postsecondary enrollment. Institutions believe that the economic downturn will resonate with the increasing demand for online courses (Allen & Seaman, 2008).

Along with an increase in demand for online learning, educational literature on e-learning findings note that learning strategies have also shifted the direction toward active and group-oriented learning (Gupta & Bostrom, 2004). This trend can also be motivated by the findings that online collaborative learning, through the use of virtual teams, has been shown to improve learning efficiency and facilitate critical thinking and communication skills (Tseng, Ku, Wang, & Sun, 2009). Corporations have a different motivation. They understand that to be successful in a highly competitive market they need the ability to quickly and economically market their products and services. These processes are now often executed by “virtual project teams” (Dustdar & Gall, 2002). VeriFone, a multinational company, has been reported to rely on teams that interact virtually to run its business (Jarvenpaa & Leidner, 1999). Microsoft also utilizes virtual teams to support major global and corporate sales and service.

Virtual teams are “groups of geographically, organizationally and/or time dispersed workers brought together by information and telecommunication technologies to accomplish one or more organizational tasks” (Tarmizi et al., 2007, page number). Virtual teams have further been categorized into a smaller subgroup called swift-starting virtual teams. Swift-starting virtual teams are the focus of this research. They have been characterized as technology intensive and primarily short-term due to the nature of technology and their rapid start-up. The challenges of working in a swift-starting virtual team can include issues with trust, communication, participation, coordination, and effectiveness (Tarmizi et al., 2007). Knowledge workers in swift-starting virtual teams need to establish a formal process and facilitate better communication among team members in order to perform work and develop clear goals and objectives. Lurey and Raisinghani (2001) emphasize that knowledge workers need to employ formally structured processes to ensure efficient and effective performance of virtual teams. Difficulties with formalizing a structured process and establishing relationships in a swift-starting virtual team are further complicated by the varying nature of collaborative tasks and the inherent inability for communication technologies to have enough depth to convey emotions necessary to establish relationships amongst team members (Warkentin & Beranek, 1999).

The ability to work in a virtual team and collaborate in distributed settings is an important and necessary skill set for today's learners to be effective when participating in collaborative learning and virtual teams. “Effective collaboration involves a set of important skills that need to be learned and cultivated, especially in the isolated environment of online learning” (Walters & Napier, 2002, p. 347).

Given the growing reliance on collaborative tasks and virtual teams, there is a need to understand what features of a potential training approach best prepares virtual learning participants in conceiving and employing structured collaboration processes and strong relational links with teammates. The training approach proposed in this research looks to provide techniques for building relational links and structuring collaborative tasks. The training approach will be evaluated through the development of process structuring and relational links in a collaborative task. The research objectives establish two key criteria: does the proposed training approach increase instances of collaboration process structuring and relational link development and is the training approach feasibility at an acceptable level. Feasibility includes usefulness, completeness, training quality, and mental effort of the training program (Kolfschoten, de Vreede, & Pietron, 2006).

The rest of this article is organized as follows. In the literature review we review related work in the areas of virtual teams, collaboration patterns, and learning theories. We present our collaboration process training approach and the research design in the study methods section. The concluding section discusses the study results, conclusions, and limitations.

In order to effectively structure and employ the unique, theoretically based training approach three key research areas were reviewed. The first research area focused on the definition of a virtual team and the inherent difficulties within these virtual teams. Difficulties include poor process structuring, lack of trust amongst teammates, poor group cohesiveness, and communication difficulties (Martins, Gibson, & Maynard, 2004). These difficulties warrant further research and the potential for a training approach based on successful collaboration techniques. Such techniques have been developed for collaboration engineering. Thus the second key research area for the proposed training approach includes an in-depth analysis of these techniques and the research that has tested these techniques. The assumption that previous training approaches relied on was that participants would learn the proposed knowledge and skills of the approach through limited contact such as hearing it described or seeing it described. In order to eliminate such an assumption from this research, the third key area explored theories and techniques for framing learning within the training approach. The end result from this research is a unique collaboration training approach for inexperienced collaboration practitioners.

While there are varied definitions for what constitutes a virtual team, most researchers agree on the following three key attributes for virtual team members: (1) members are responsible for individual tasks guided by a common purpose, (2) members must rely on some form of communication technology more than face-to-face interaction, and (3) members are likely geographically dispersed from each other (Schiller & Mandviwalla, 2007). For the purpose of this research, we adhere to the following definition by Maznevski and Chudoba (2000) that captures these attributes. A virtual team is defined as “a group of people who interact through interdependent tasks guided by common purpose and work across, space, time and organizational boundaries with links strengthened by webs of communication technologies” (Maznevski & Chudoba, 2000, p. 473). There are multiple factors within a virtual team environment that inherently impact the collaborative success of virtual teams. These factors include time differences, mismatches in expectations, cultural differences, different levels of experience, and a lack of norms for communication (Munkvold & Zigurs, 2007). The direct result of these factors can be poor process structuring, lack of trust amongst teammates, and ineffective communication. The end result is that many virtual teams do not develop a shared understanding of what needs to be done to achieve their goal. Understanding ways in which these barriers to quality virtual collaboration can be addressed calls for further research. We begin by focusing on some of the most frequently used theories in virtual team research and then defining the variables impacting the success and performance of virtual teams.

Virtual team research has several key theories that focus on the interplay between the use of collaboration technologies and relationship development in virtual teams. Reliable research indicates that the two most frequently used theories in virtual team research are adaptive structuration theory and the media richness theory in virtual team research. Of adaptive structuration theory Schiller and Mandiviwalla (2007) “emphasize the interplay between technology and the social process of technology use, illuminating how multiple outcomes can result from implementation of the same technology” (p. 18). Media richness theory (MRT) is primarily concerned with media preferences and usage in organizational settings. MRT suggests that communication media can be ranked on a richness continuum based on their ability to handle equivocality and uncertainty (El-Shinnawy & Markus, 1997). The development of relationships and trust between team members is directly and negatively impacted by the sole reliance on computer-mediation. MRT and social presence theory state that computer-mediated group interactions can be characterized by their “inability to share socio-emotional information and cues needed to develop trust, warmth and other interpersonal affections” (Warkentin & Beranek, 1999, p. 274). Developing relational links consists of performing activities that relate to the wellbeing of the group and individual members. Relational link development fosters and maintains the occurrence of trust in virtual teams. Relational links can be developed through such steps as defining member roles and establishing consistent patterns of communication (Warkentin & Beranek, 1999). In face-to-face groups the development of relational links are quickly and easily established through facial expressions, tone of voice, and other nonverbal cues. These cues can quickly stimulate conversation, convey meaning, and drive agendas. Establishing relationships within virtual team members has proven to be more challenging than in the face-to-face environment (Warkentin & Beranek, 1999). This challenge often occurs because ad-hoc and swift-starting groups do not have time to develop relational associations. Team members will also often focus on task activities at the expense of relational link development (Munkvold & Zigurs, 2007). Why is the establishment of relational links and trust important? The importance of these two factors directly relates to their impact on virtual team processes. Trust can increase confidence and security within team member relationships and encourage an environment in which information can be open and freely exchanged (Jarvenpaa, Knoll, & Leidner, 1998). The inability for virtual teams to freely exchange information can, and more than likely will, negatively impact team performance. A study by Iacono and Weisband (1997) found that trust development in virtual teams was one of the key predictors toward higher performing teams. Higher performing teams were more efficient, able to meet deadlines on time, and were more socially present for each other.

Another conception of virtual teams focuses less on the development of trust and relationships and more on the processes utilized by these teams and the impact this process has on collaborative success. After studying widely used theories in virtual team research, Schiller and Mandviwalla (2007) divide virtual team research into three main categories: team inputs, team processes, and team outcomes. This distinction is important in that these key research areas reflect the characteristics of virtual teams that directly correlate with outcomes and the development of shared understanding amongst team members. Another key study developed similar categories. Lurey et al. assessed virtual team effectiveness. They found that are three core criterion in this framework: team performance, work process and individual team member satisfaction (Lurey & Raisinghani, 2001). These studies show that in order to evaluate virtual team performance it is important to assess the impact of work processes on the outcomes of collaboration. In each of these studies work processes showed a correlative relationship to team performance out comes and team member satisfaction. Work processes are the structural elements utilized within virtual teams to complete tasks. Work processes can include process development and task structure (Munkvold & Zigurs, 2007). Due to the difficult nature of virtual teams, these teams require more structure to perform their work (Lurey & Raisinghani, 2001). The process structure activities within a virtual collaboration can be hindered for several different reasons. First, those individuals who are comfortable with a virtual team environment may experience difficulties due to the differences in experience levels of participants (Munkvold & Zigurs, 2007). Second, there are currently a number of different collaboration tools utilized which vary in complexity (Munkvold & Zigurs, 2007). Lastly, the complexity of group process and the lack of knowledge of how to structure group processes is prohibitive to virtual group collaboration (de Vreede & Briggs, 2005; Munkvold & Zigurs, 2007). The frequent assumption is that individuals within a group have the inherent ability and skills necessary to work as a group to structure tasks and develop processes toward successful completion of a group goal.

The development of relational links and process structuring within virtual teams have been shown to have an impact on the work processes of a collaborative activity. The previous discussion addresses techniques for the development of relational links, such as the establishment of communication norms, which is the first contribution of this research. The following discussion will explore a promising facilitation technique for establishing the development of process structuring, collaboration engineering (CE).

The skills necessary to properly facilitate a collaborative activity are not inherent to most individuals. These skills are crucial because many collaboration activities can prove to be successful when facilitated properly. At the height of group decision support (GSS) popularity, professional facilitators were extensively trained and utilized to guide novice groups through collaborative activities. Learning how to manage a typical GSS system was thought to take a few days, whereas it was thought to take a year or more to truly understand how the features could be utilized effectively in the service of group dynamics (Briggs, de Vreede, & Nunamaker, 2003). Professional facilitators are found to move a group through a collaboration process more efficiently and effectively than if a group is left to its own devices (de Vreede & Briggs, 2005). The result of this face-to-face facilitation was a decrease in project completion time and improved results (de Vreede & Briggs, 2005). The difficulty with professional facilitation utilization is that it is expensive, difficult to maintain, and difficult to find (Briggs et al., 2003). Collaboration engineering was developed as an answer to the decrease in professional facilitation utilization and GSS system utilization. Organizations recognize the benefit of facilitation, but need a method to improve its feasibility. Collaboration engineering seeks to provide organizations with the benefits of professional facilitation through the use of available resources. “Collaboration engineering is an approach to the design of reusable collaboration processes and technologies” (de Vreede & Briggs, 2005, p. 1).

Collaboration engineers support group processes and structure for reoccurring activities. Once engineers have designed a collaborative environment, they are no longer a part of the collaborative process. “In collaboration engineering the collaboration engineer designs a reusable and predictable collaboration process prescription for a recurring task, and transfers the prescription to practitioners to execute without the intervention of group process professionals” (Kolfschoten, 2007, p. 95). The belief behind this strategy is that the practitioner does not need to have an understanding of the process structure or facilitation techniques. This belief is problematic for several reasons. First, not all collaborative activities will have access to structured patterns created by a collaboration engineer. Second, collaboration engineering often only focuses on reoccurring collaborative tasks. Thus excluding the many instances of collaborative tasks that are not deemed reoccurring, but which still need structure. In the virtual team environment the majority of tasks can be deemed ad hoc. This limitation is unnecessary as there is enough depth within the collaboration engineering process to adjust the prescription for ad-hoc teams.

Within the process prescription created by a collaboration engineer is a set of specialized and scripted patterns of collaboration. These patterns, called thinkLets, were developed in order to achieve collaboration engineering goals. The concept of thinkLets was introduced to reduce the difficulty in which practitioners found when trying to facilitate a process prescription developed by a collaboration engineer. The general process framework for the application of thinkLets was first prescribed by Lowry and Nunamaker (2002) in their work with collaborative writing. A thinkLet is a way to create a pattern of collaboration that contains building blocks for group processes. Essentially, thinkLets are packaged, repeatable, and transferable facilitation techniques that can be deployed to create predictable patterns of collaboration among a group of people with a shared goal, during a collaborative process (de Vreede, Kolfschoten, & Briggs, 2006). Each thinkLet supports one or more of the six general descriptive patterns of thinking in performing an intellectual task collaboratively, namely generate, reduce, clarify, organize, evaluate, and build consensus (Tarmizi et al., 2007). Like design patterns, thinkLets serve multiple purposes in the design and deployment of collaboration processes (de Vreede et al., 2006). Thinklets have primarily been used to design collaboration processes to enable process structuring by collaboration engineers, not practitioners or participants. “Thinklets offer collaboration process designers building blocks to assemble a collaborative work practice. They offer an overview of capabilities required to support collaboration” (Kolfschoten, 2007, p. 97).

Efforts have been made to scaffold the design process for collaboration engineers. The first study, completed in 2007, focuses on developing a collaboration engineering design approach that provides support for novice collaboration engineers, creates insight into the steps within the collaboration process, provides a starting point for creation of design support tools, and provides a basis for the training of collaboration engineers (Kolfschoten & de Vreede, 2007). The novice collaboration engineer received the completed process design from the collaboration engineer design approach developers and very limited training on how to facilitate the process. Their findings indicate that although the approach and supplemental material was useful, it was difficult for students to learn and apply the complex elements. The process design approach in this study was instrumental to the development of the proposed training approach in this research in that it provides the fundamental framework for the structuring of processes.

A second study also looks to evaluate the feasibility of collaboration engineering in a different manner, through a distributed environment. Two important concepts in this study, which are relevant to and difficult in virtual teams, are the “achievement of shared understanding through process use and team leadership” (Tarmizi et al., 2007, page number). This study was one of the first studies to address this application. This study was also unique in that it paid specific, yet limited, attention to relational link development. “Although existing CE techniques primarily focus on task-related processes, these techniques can also be specifically designed to promote relational development” (Tarmizi et al., 2007, page number). Results of the study showed that virtual teams still found it challenging to take advantage of the proposed collaborative process pattern concepts and techniques to improve various levels of satisfaction within a virtual team. These results indicate the need for further research in the area. “For virtual team members, we emphasize the need for training and continual reinforcement of the types of processes that are most helpful for different tasks” (Tarmizi et al., 2007, p. 93). The researchers offer the suggestion of “pretraining” on virtual team membership with the end effect of possibly encouraging team members to think differently about virtually collaborating and to teach them specific things to which they need to pay attention to.

The field of collaboration engineering contributes a number of techniques and processes that have the potential to provide novice practitioners with the tools to effectively collaborate in a virtual team. The research has shown that these virtual team collaboration techniques, in combination with the development of relational links, are crucial for groups to establish collaborative success. This supports the need for a similar design process and training approach for practitioners. We argue that providing training about process structure and relational link development to practitioners can better prepare them to engage in collaborative activities in a structured manner. Our proposed approach will be unique in that it will focus on users with limited knowledge of creating structured processes for collaboration activities.

In order to design an effective training approach it was important to align the training objectives of the approach with a structured instructional design process. For this research the training objectives were based on Bloom's Taxonomy.

Learning has been described as “a change of state of a human being that is remembered and makes possible a corresponding change in the individual's behavior in a given type of situation” (Gagne, 1984, p. 377). Because the facilitation and training protocol is premised on participant learning, it makes sense to anchor the objectives of the approach in learning theory. Distance learning theorists generally define learning in terms of its socioconstructivist tendencies that acknowledge the roles of discourse, dialogue and modeled behavior in learning. While socioconstructivism remains the basis for the interactions described in the collaborative facilitation of this study, a more structured framework is needed to support the objectives of participant learning. For this task the authors have applied Blooms taxonomy of the cognitive domain.

Bloom's Taxonomy of the Cognitive Domain was constructed in 1956 as a collaborative effort between Benjamin Bloom and a group of educational psychologists at the University of Chicago. Bloom's Taxonomy is one of the most universally accepted approaches to understanding the nature of learning outcomes. Because the taxonomy is constructed on a hierarchical basis, it is possible to identify the complexity of a learning outcome given the stated instructional objectives as well as the assessed outcomes (Bloom, Engelhart, Furst, Hill, & Krathwohl, 1956). Traditional uses of Bloom's Taxonomy focus on it as a benchmark for measuring student's level of understanding of a subject. In this research the framework was alternatively used as a way to structure and facilitate the objectives of the training program in a hierarchal manner. The learning objectives serve as a guide within the training approach to move students from the remembering level of mastery to the evaluating and creating level of mastery. Bloom's Taxonomy was not utilized as a basis for assessing learning outcomes of the collaborative activity itself.

The focus of the literature review section was to develop a fundamental theory for the framework of the proposed virtual team collaboration training approach. Following the literature review a comprehensive training approach was created and tested with the goal of establishing relational link and process structuring development within a virtual team.

While the literature suggests there are a number of approaches to facilitate learning and collaborative work in virtual teams, the literature is scant in providing outcomes of collaboration engineering in the distributed environment. This study reports the outcomes of a virtual team collaboration (VTC) training approach in select undergraduate courses of a technology intensive, small Midwest university. This section includes an in-depth discussion of the training approach goals, structure and theoretical basis behind the structure. The VTC training approach is designed with the main goal of fostering the development of collaboration process structuring skills in participants with no previous formal training. The secondary, yet essential goal of the training approach is to provide participants with key concepts and ideas for enhancing relational links along with other team members. Toward this end, there are two key components in the training approach. The first component focuses on the development of relational links among virtual team members, while the second component focuses on providing structure to collaborative work processes.

The VTC training approach contains six sequential individual modules. Each module within the VTC training approach has a foundational theory. The first module, Module 1: Orientation and Trust Building, leverages The Team Performance Model (TPM), proposed by Drexler, Sibbet, and Forrester (1988) as the foundational theory for the design of training activities focused on relational link development. The first two stages within the TPM—orientation and trust building—were used because they focus solely on the development of relational links. Module 1 also includes example orientation and trust building activities such as icebreakers, group formation activities, and the rules of netiquette. These activities encourage the establishment of communication norms and mutual regard for teammates. Modules two through six of the VTC training approach leverage process structuring techniques using the (CE) process design approach and provided example application of the material. Two adjustments were made to the CE process design approach. First, the design approach has limited support for relational link development, thus the addition of the TPM model for Module 1. Second, the CE process design approach is an extremely in-depth approach for process structuring. In a study by Tarmizi et al. 2007, which utilized the CE design approach in a distributed environment, they found students were often overwhelmed by the process and the material. To help alleviate some of these issues, a simplified version of the CE design approach components was developed for the proposed training approach. Table 1 provides a representation of the VTC training approach framework.

The VTC training approach leverages Bloom's Taxonomy of Learning as benchmarks for developing the learning objectives of each module. There were six corresponding lesson objectives and activities requiring completion for each module.

To engage learners in the training approach, and support their participation, Desire2Learn (D2L), a learning management system (LMS) was used. Any LMS can be used to administer the training approach. The LMS will need an area to post content and links and a tool for questioning students, such as a quiz or survey tool.

In order to evaluate the impact of the training approach this study focuses on the feasibility of the training approach and the relationship between collaboration training and work processes. Figure 1 visually diagrams the theoretical framework.

The research objectives for the study are to establish two key criteria. First, does the VTC training approach increase instances of collaboration process structuring and relational link development? Second, is the VTC training approach feasibility at an acceptable level? Feasibility includes usefulness, completeness, training quality and mental effort of the VCT training approach (Kolfschoten et al., 2006).

The study for VTC training occurred in the fall of 2009, in an introductory graphic design history course delivered exclusively asynchronously and online. The students in the course were primarily sophomores and juniors. The majority of students in the course had taken one or two online courses prior to enrolling in this course. The university in which the course was taught was a small, public institution that focuses on information and educational technology. In order to control any factors that may influence the outcome of the study, there were two projects for the course, Project 1 and Project 2. There were also two phases to the experiment that coincide with the development of the two projects in the course. Each project had a similar deliverable but focused on a different content area. Care was taken to make sure that the requirements for the project were indeed collaborative and not merely cooperative in nature. The difference being that collaboration involves all participants at each stage of project development, while cooperation may only involve the parsing of tasks to team members to complete individually.

The study applied an experimental/control group approach by administered two separate surveys for data collection via the course management system. The first survey was the Virtual Team Survey. The Virtual Team Survey is a survey questionnaire adapted from Lurey and Raisinghani (2001) and composed of specifically created relational link development questions. The survey also includes several questions on process structuring in groups. The Group Training survey is a survey questionnaire adapted from Kolfschoten et al. (2006) that was administered to test the feasibility of the VTC training approach. Specifically, this survey evaluates the usefulness, completeness, training quality, and mental effort of the VTC training approach.

For Phase I, Project 1, the students were randomly assigned to four groups of 4 and one group of 3. Students were then given instructions to collaboratively complete the project. Students did not receive any collaboration related training. They were provided the instructions for completing the course project and were required to complete the project using the methods and the tools they develop. Each group was also required to keep track of and report all group related activity. This included meeting agendas and all communication such as e-mails, chats, and online discussions. Upon completion of Project 1, the Virtual Team Survey was administered to each individual in the class.

The second phase began with students individually completing each module of the VTC training approach. The approach was given for credit only. Upon completion of the VTC training approach, the Group Training survey was administered to test the feasibility of the approach. Students were again randomized into different groups and asked to complete Project 2. Each group was required to keep track of and report all group related activity. Upon completion of the project, the Virtual Team Survey was again administered to each individual in the class. See Figure 2 for an overview of each phase of the experiment. In order to evaluate the survey data, a one-way (ANOVA test was used to test the mean variation, and factor analysis will specifically evaluate data correlations.

The first research objective focused on discovering if the VTC training approach increased instances of collaboration process structuring and relational link development. The statistical analysis included two key tests. First, a one-way ANOVA test was used to test the means of several populations. Second, factor analysis was used to determine if underlying factors were responsible for the correlations in the data.

The one-way ANOVA test was run on the results of the Virtual Team Survey given in the study. Tukey's family error rate was set to 5. A significance level of .05 was used for all statistical analyses. With this test we were looking to see if there was a statistical significance between means. From the results we evaluated each p value for its significance, confidence level and looked to see if Tukey's test contained 0. Within the survey there were a total of 28 questions, split into two different sections. The first 19 questions were focused on relational link development and the last 8 questions were focused on process structuring development. In the first 19 questions the pvalue was considered significant in 8 out of the 19 questions. In those 8 questions, 4 questions exhibited a significant p value, a nonoverlapping 95% confidence level and Tukey's test did not contain a 0. See Table 3.

In the last eight questions, which focused on process structuring, the p value was considered significant in two of the eight questions. These two questions also exhibited a significant p value, a nonoverlapping 95% confidence level and Tukey's test did not contain a 0. See Table 3.

The results of the ANOVA statistical analysis show that there was evidence of statistical significance between the population means. If there were no significances indicated from the test, the result could be a null hypothesis.

The next step in the statistical analysis was to run a factor analysis on the survey to determine if the results would show that there were two factors. Factor analysis was first run using a varimax rotation. The results of this first analysis indicated that all of the questions in the survey instrument were loading on the same factor, instead of two factors. Questions 1-19 should have all loaded on Factor 1 and questions 20-28 should have loaded on Factor 2. Factor 1 explained 83% of the variance between questions and Factor 2 explained less than 1%. In the varimax rotation, the overall MSA is less than .5. Measure of sampling adequacy (MSA) values must exceed .50 for both the overall test and each individual variable. Variables with values less than .50 should be omitted from the factor analysis one at a time, with the smallest one being omitted each time. These results were not the results that we were specifically looking for. One issue with the use of the varimax rotation is that it necessitates that there is no correlation with the factors. Within this study there may be some overlap.

The second research question was to determine if the VTC training approach feasibility was at an acceptable level. Feasibility included usefulness, completeness, training quality, and mental effort of the VCT training approach (Kolfschoten et al., 2006). To measure the constructs a survey was given to all training approach participants. The first section of the survey utilized a 5-point scale, 1 (not at all useful), 2 (somewhat useful), 3 (neutral), 4 (useful), 5 (is very useful). This section evaluated the usefulness of the training approach steps. The results are in Table 4.

The usefulness of the training approach itself and it application was then evaluated. This section of the survey utilized a 5-point scale: 1 (very much disagree), 2 (disagree), 3 (neutral), 4 (agree), 5 (very much agree). The results are in Table 5.

To evaluate the completeness of the approach we asked if the training materials were complete and also provided a text box for individuals to have the opportunity to discuss materials that they felt were unnecessary. Additional comments were not added by participants about unnecessary material. The results are in Table 6.

We then measured the training quality of different elements within the training approach (see Table 7).

The last construct evaluated was mental effort. See Table 8 for the results.

The purpose of this study was twofold. The first research objective was to determine the impact of the virtual team collaboration (VTC) training approach on the development of process structuring and relational links in a virtual team. Past research in virtual teams indicates that a formal process to perform work, develop clear goals and objectives, and facilitate better communication among team members needs to be established to ensure efficient and effective performance of virtual teams (Lurey & Raisinghani, 2001). We hypothesized that the proposed training approach would cause an increase in the instances of process structuring and relational link development. The ANOVA comparisons demonstrated significance and supported the research hypothesis. These findings suggest that the proposed training approach increases instances of process structuring and relational link development in virtual teams.

The second research object was to evaluate the feasibility of the training approach. Feasibility included usefulness, completeness, training quality, and mental effort of the VTC training approach. Feasibility needs to be at a certain level in order to make sure that the participants can successfully complete the approach and to assure that the training approach can be administered in different settings by different individuals as needed. The first construct evaluated was usefulness. Overall the majority of participants felt that that each step within the training approach was useful. The one item within this category that was not deemed useful was the lecture videos. These videos were a duplication of the OneNote and PDF material to account for different learning styles. It would be possible to make them an optional part of the training approach. Participants also felt that they will use these techniques and felt better equipped to work in a group. Training quality had somewhat mixed results. Participants felt the material was not well explained, but that the material was presented in a logical manner. The training approach itself did not include an introductory module; this was implied as part of the responsibility of the individual administering the approach. An introductory module could easily be included to describe the goals of the training approach and include directions. The last construct, mental effort, also showed mixed results. Participants reported that they did feel the training approach required a lot of mental effort and that it was tiring, but they did not all agree on the idea that it was difficult. In this instance students were given a week to complete the training approach, along with additional required work for the course. If participants were not required to complete dual deliverables, this may have positively impacted the results. Overall the participants responded with either neutral or slightly positive answers to the majority of the questions relating to the feasibility of the training approach. Slight adjustments can easily be made to improve these findings.

We must note that there are limitations to the results found in this study. The results of the factor analysis indicate that some changes need to be made in order to improve the Virtual Team Survey. This instrumentation focused solely on process structure and relational link development. In order to improve the results, each question in the survey instrument needs to be evaluated. The results of the factor analysis indicate that the questions are not clearly loading on two factors, when they should be. Thus the questions on the survey that specifically pertain to these factors need to be evaluated, reworked, and eliminated as necessary. The survey instrument as it is now has 19 relational link questions and eight process structuring questions. The number of relational link questions could be reduced significantly and the process structuring questions would need to be revised. The data gathered from the Virtual Team Survey are also self-reported which can result in potentially biased results. Factors not under the control of the researchers include the inability for all subjects to complete all the appropriate components of the training approach. Subjects who did not complete all required aspects of the training approach were not included in the study in two ways. First, their incomplete submissions were eliminated. Second, when putting together the groups for collaborative activities, care was taken to make sure those subjects who had completed the required training were put together in groups.

Further research opportunities exist based on the outcomes of this study. First, this study focused on the increased instances of process structuring and relational link development in a collaborative task. The next step would be to further evaluate the increased instances of process structuring and relational link development on collaborative success. Additional evaluative measures could be taken to include an analysis of the process outcomes and quality of collaboration as evidence by the collaborative discourse produced by student.

The use of collaboration in a distributed setting has been in place for many years. Throughout these years there have been a number of theories developed as to what causes some teams to perform at a higher level than others. The training approach offered in this research builds on those theories to create a precollaborative activity to cultivate and teach a set of important skills for knowledge workers to be successful in virtual teams.