Literary Discussion from the Work of Kaldis, Koukoravas, and Tjortjis

This journal article gives the thought process, structure, hypotheses, implementation and results of a study that was done with a group of 204-second-year undergraduate IT students who were grouped into teams of four to six students each. Given the option to choose their own groups, with the remaining students being assigned together at random, thirty-three of the teams were self-nominated and the remaining nine were randomly formed. Those conducting the study were looking to identify the problems existing in academic work done in teams, and introduce various tools that could help improve upon those things identified. They decided to implement a Network Organizational Structure, to decentralize authority and encourage equal contributions from members based on individual skills and talents, and they decided to use IT to improve communication problems that normally arise in a team-oriented environment.

In introducing IT to aid in communication challenges that many times exist within a group setting, the experimenters decided to use Groupware, a software tool that allows for instantaneous communication between group members, regardless of their location. This worked well because it allowed all members of the groups to communicate with one another during their most convenient timeframes, encouraged everyone to be an active part of the process, and distributed the power among the group equally. It also was an added plus in teamwork as it instantly stripped any preconceptions members might have formed against each other based on non-issues such as ethnicity, gender, age, or social standing.

The experimenters also decided, based on collective research, that peer evaluation is an important technique for allowing for the most effective structure one can have in teamwork, and as such they introduced software to provide for it. This worked well because traditionally speaking, the time it requires to track, assess, and distribute peer reviewing for faculty can be a huge burden. With the software doing the work, the students were able to contribute anonymously, and the data was able to be organized and distributed very quickly and in a variety of ways. This worked well because it was a benefit to the students as they received their feedback, was a bonus for the faculty members because it reduced the workload tremendously that would have been needed for such a task, and it enhanced motivation within groups as each member knew the peer evaluation was to be a part of their grade.

Graphs were included in the article to show the flowchart process of various teams including those without a structure, and those with a hierarchical structure, a core-periphery structure, one with structural holes, those working towards achieving a network organizational structure, and finally that of a network organizational structure.

In going forth with the new Groupware, the structure of the experiment was formed. As described in the article, a Network Organizational Structure is, "where independent people act as independent nodes to work together for a common purpose. A network comprises node and connections-links" To paraphrase, the nodes in this case would be the team members. The links would be the interactions taking place amongst and between the nodes. This was the type of organization they felt would be most productive in enhancing teamwork situations in upper education.

With that in mind, the conductors came up with the following five hypotheses to test:

1). Individual difference in contribution among team members are significant

2). The information of a network structure is correlated with team performance

3). Team member performance is correlated with the team's structure.

4). Groups of good students perform better than mixed groups or groups of mediocre students.

5). Team self-nomination affects team performance negatively

The experiment was conducted and a peer evaluation was designed using a rating scale for thoroughness. Student contributions were anonymous and comments were for the professor's eyes only so as to not to cause any discordance amongst team members. The software was used to give analysis of each team collectively and for individual members, to provide self-evaluation, to assess objectivity of the contributor, and to allow student access to the scores after the course was completed for self-evaluation. It was also set up for the faculty to be able to see the rankings of the teams and team members.

The results are summarized as follows:

Hypothesis Number One: The software implemented proved this theory true as it calculated student participation and noted that, as speculated, team members did not always contribute equally. Deviations were low among some members and very high when comparing others. The conductors wondered aloud if based on this proven hypothesis, grades should be allocated with emphasis on the group activity.

Hypothesis Number Two: Results here showed that there was almost no correlation between group performance and how leadership varied among the team members. With the Mann-Whitney test in place to determine the mean leadership variances from the subset of the best and worst team members performance, it was agreed that "the formation of a network structure is correlated with team performance" could not be accepted as true. It was also noted, however, that these results did not mean network structure did not improve group performance, only that "group performance may or may not improve when a network structure is adopted."

Hypothesis Number Three: The Spearman correlation coefficient was applied here between the structure of each team and student quality, and results found that students who did receive high marks individually did tend to work in a network structure and those who did not receive individual high marks tended to work in different structure types, mainly hierarchical. It was noted, however, that outside influences must also be considered in how teams and their individual members' work is conducted and affected.

Hypothesis Number Four: The top 20% of group performance and the bottom 20% were identified here as the subset, and with the Mann-Whitney test in place, it was determined the theory that 'groups of good students perform better than mixed groups or groups of mediocre students' could not be accepted. This was the notably the most surprising result to the testers. It was noted, however, that this result could indicate that when a group is working well, standing out as an individual may simply be harder than if the group were performing poorly.

Hypothesis Number Five: The results of this theory disproved that self-nomination affects team performance negatively. Both the Mann-Whitney and Spearman correlation coefficient were applied and in each situation it was proved that when team members are of a group's own choosing, they are apt to perform better than if the team were not self-nominated.

The most important thing this experiment revealed, however, was that IT is a huge enhancement for implementing a network organizational structure among academic teaming. Among other things, it allows for better cohesion, coordination, and communication between individual members and between groups; leading ultimately to a better overall experience for students in higher education who are assigned to team work.

Kaldis, E., Koukoravas, K., & Tjortjis, C. (2007, September). Reengineering Academic Teams Toward a Network Organizational Structure.

Decision Sciences Journal of Innovative Education, 5(2), 245-266. Retrieved September 11, 2008, doi:10.1111/j.1540-4609.2007.00140.x