In gamification, abstract points or achievement levels are often used as rewards (Groh, 2012), but simply applying points or badges does not always make an effective gamification design (Falkner and Falkner, 2014). There is potential to go beyond that in the design of gamification. Publication V presented a new method for the design of adaptive gamification, where the discovered user profiles and Bartle’s (1996) theory of player types can be used to choose a suitable gamification approach for each type of user.
76
4 System for increasing computer-supported collaboration in engineering teamwork
Table 4.5: Comparing the averages of iteration evaluation results
Survey Question Iteration
1
Iteration 2 How useful did you find the following features of the system?
1: Asking questions <> 2: Reporting issues 4.02 4.16 1: Answering questions <> 2: Being able to offer help 3.9 4.27 1: Being able to vote <> 2: Being able to /like actions 2.82 3.55 1: Earning scores and badges <> 2: The point system 2.1 3.33 1: Seeking for useful conversations <> 2: The chat
functionality
4.26 4.27
The system in all 3.67 4.16
As future work a third design iteration could be created, improving on the gamification aspect of iteration 2, by utilizing a profile-based adaptive gamification approach. For example, Bartle (1996) defines the explorer type of a player, who enjoys discovering new knowledge in the system, like new areas. An explorer would not care for challenges that provide an opportunity to earn more points, but would enjoy challenges that involve discovery. To continue the example, in Publication V a social observer type of a profile was discovered, and the interaction analysis revealed that the students in this profile discussed often with the members of other teams, but did not always share the knowledge about the other teams back to their own team. With a targeted gamification approach, this social exploring activity could be harnessed by prompting social users to contribute information back to their own team, and acknowledging this activity as a productive part of collaboration.
This chapter presents a summary of the results of the individual studies and discusses their theoretical and practical implications. It also relates the results to earlier research and existing theories.
5.1 Practical results and implications
The foremost practical contribution of this study is the research artifact itself, its design, and the evaluation results of the test cases. These cases showed that collaboration between project teams of software engineering students can be increased by publishing issues and facilitating explicit goal management. It was also shown that students’ use of the system can be increased with gamification, with positive results in user satisfaction and the frequency of system use.
The research project advanced the state of computer-supporting collaborative learning by creating the research artifact described in chapter 4. The artifact has been publicized in scientific and professional communication venues, with plans to publish the ongoing software project. The course design of computer-supported collaborative courses was developed further when adapting them to utilizing the research artifact. The course design for an intensive-format software engineering course (originally introduced by Porras et al. (2007) and Alaoutinen et al.
(2012)) was improved by adding computer-facilitated communication and aspects of inter-team collaboration. From a design science perspective the prototype instantiation of each iteration can be considered instantiation artifacts and one of the practical results of the research programme.
The artifact design took account of the following factors that had the potential to increase collaboration: flexible and issue-based communication, inter-team collaboration, shared goal setting, and explicit goal and contribution visibility.
By leveraging these methods, software engineering students can be encouraged to collaborate online with the application of gamification, and this collaboration has the potential to improve the outcomes of both learning and software quality.
These results were achieved with the first-iteration research artifact and the related test case. In the second iteration test cycle, the gamified collaboration system increased the amount of inter-team collaboration and shared problem solving in a software engineering project course.
Several project-based collaboration platforms exist, such as Redmine5, that support task negotiation and discussion between team members. However, the system presented in this study concentrates also on a low barrier of entry, emphasizing a low effort required to initiate communication, and provides positive feedback
5http://www.redmine.org/
78 5 Discussion
for actions. Even teams that cooperated in the beginning, or split tasks rigidly and started to work on them separately, often started to collaborate with each other or members of other teams when they noticed that they were working on similar issues. In many cases the collaboration also extended to collaborative learning, where the team members shared knowledge in addition to solving the programming issue.
Collaboration has been defined to be symmetrical in action, knowledge and status (Dillenbourg, 1999b). Many collaboration tools try to increase positive interdependence between project members by creating a hierarchy that promotes interdependencies and clearly perceived individual accountability. However, this undermines symmetry in action and status, especially if only a part of the project adopts the use of the system, and the hierarchy can cause responsibilities to break down into separate niches. The system presented here creates flexible avenues for project members to contribute and to help project members to reach out to help each other. This problem solving -related collaboration enables the participants to learn from each other, promoting the learning goals further. Acknowledging the help, publicizing the solved goal, making the progress of the team explicit, and promoting collaborative communication with gamification promotes positive interdependencies without having preset hierarchies.