Collaborative Game-based Learning - Automatized Adaptation Mechanics for Game-based Collaborative Learning using Game Mastering Concepts

Learning and playing represent two core aspects of the information and communication society nowadays. Both issues are subsumed in Digital Education Games, one major field of Serious Games. Serious Games combine concepts of gaming with a broad range of application fields: among others, educational sectors and training or health and sports, but also marketing, advertisement, political education, and other societally relevant areas such as climate, energy, and safety. This work focuses on collaborative learning games, which are Digital Educational Games that combine concepts from collaborative learning with game concepts and technology. Although Digital Educational Games represent a promising addition to existing learning and teaching methods, there are different challenges opposing their application. The tension between a game that is supposed to be fun and the facilitation of serious content constitutes a central challenge to game design. The often high technical complexity and especially the instructors' lack of control over the game represent further challenges. Beyond that, the distinct heterogeneity of learners who often have different play styles, states of knowledge, learning speed, and soft skills, such as teamwork or communication skills, forms a pivotal problem. Apart from that, the vital role of the instructor needs to be taken into account. Within the scope of this dissertation, the problems mentioned above are analyzed, concepts to solve them introduced, and methods developed to address them. The first major contribution contains the conceptualization of a framework for adaptation of collaborative multiplayer games as well as for the control of those games at run-time through an instructor using the Game Master principle. The core concept hereby addresses the design of a model to represent heterogeneous groups and to represent collaborative Serious Games. Based on that, a novel concept for adaptation of collaborative multiplayer games is developed, implemented, and evaluated. Automatic recognition and interpretation of game situations, as well as determination of the most well suited adaptation based on the recognized situations, is a major challenge here. Further, a concept is developed to integrate an instructor in a meaningful way into the course of the game, giving him/her the necessary resources to recognize problems and to intervene and adapt the game at run-time. Therefore, it will be taken into account that the elaborated concepts are applicable in a generic way independent of the underlying game. The second major contribution of this work is the conceptualization and design of a simulation of players and learners in a collaborative multiplayer game that behave realistically based on a player, learner, and interaction model. This is supposed to enable an evaluation of the adaptation and Game Mastering concepts using freely configurable player and learner types. The concepts introduced and developed within this thesis have been thoroughly evaluated using a twofold approach. As a test environment, a collaborative multiplayer Serious Game was designed and implemented. Within that simulation environment, the developed Game Mastering and adaptation concepts were assessed and tested with large sets of virtual learners. Additionally, the concepts were evaluated with real users. Therefore, two different evaluation studies with a total of 60 participants were conducted. The results of the conducted evaluations help to broaden the areas of application of Serious Games as well as to improve their applicability, hence raising acceptance among instructors. The models, architectures, and software solutions developed within this thesis thus build a foundation for further research of multiplayer Serious Games