The research methodology of APOGEE will follow an iterative and incremental approach extending with learning by doing the methodology defined by Kirjavainen especially for generation of game concept and game design and development folliwing a multi-disciplinary prescriptive development model, using an iterative and incremental approach and agile techniques such as the SCRUM process control methodology. As most appropriate for the project as far as it covers all stages of the project research and development process and is combined with learning by doing; with phases as follows:
- Identification of research problems and formulation of research questions and tasks. Here, learning by doing addresses an increased exchange of scientific and technology solutions ideas between different domain specialists such as cultural heritage specialists, game designers, platform and game developers, and GBL practitioners and pedagogues. Phase objectives: adequate and precise description of the research context (problems in formal game description and semantic content organization, generation of video games, game scenarios/didactic methods/tasks for adaptive games) and detailed research formulation - research questions and tasks, identification of needs of target users of the platform (educationalists and teachers) and of the educational video games constructed by the platform (5-8 class school students).
Answering the research questions - by theoretical and practical work with learning by doing.
- Definition of formal language for description of hierarchical mazes, incl. semantically structured learning content
- Design and development of an innovative open software platform based on an existing game engine (such as Unity 3D ) and used for generation of customized hierarchical 3D video maze games with simple and intelligent virtual players, whereupon the construction is based on declarative game description and semantically structured cultural artifacts, constructed by means of a metadata-driven visual maze editor.
Practical proving the answers to the research questions. Phase objectives: implementation of software prototypes of player-centric video games using the generation platform from the previous stage. The iterative and agile (SCRUM-based) risk-driven process will consist of two proof-of-concept iterations, namely:
- Development of non-adaptive prototypes of desktop and online video games for Bulgarian medieval history, with simple virtual players – including proactive game scenarios, tasks and 3D model content authoring. The game-based approach uses modern learning paradigms by developing a series of interconnected activities in which learners are not only active participants but also have the freedom to choose their own learning pathway in the game.
- Development of adaptive prototypes of smart adaptive video games for history education, and questionnaires for online surveys. All the experiments will be conducted online in order to include recognition of player affect inferred EDA and BVP dynamically measured by custom measuring device and the RAGE game asset form measuring the player affect , and recognition of emotions based at least on facial expressions by means of software libraries like Affdex SDK.
Practical experiments with the target groups – includes testing of the game prototypes of the chosen domain in real conditions, with online experimental setup including consent form and information sheets, smart adaptive video games for history education with simple and intelligent virtual players, and questionnaires for online surveys. All the experiments will be conducted online in order to include more participants for obtaining statistically significant results. Phase objectives: definition and execution of online experimental studies, including game sessions with non-adaptive and adaptive versions of the games and pre- and post-game questionnaires, addressing previous gaming experience, game engagement and experience , and generation platform usability.
Analyses and validation of the results of the study. Phase objectives: analysis and validation of experimental results and their reflection on final games’ prototypes, by means of analysis of variance (ANOVA), correlation analyses, and principal components analysis and/or factor analysis – all applied for validating (1) platform usability for the target user group (teachers and educationalists), (2) learning analytics of adaptive and agent-faciletated GBL for school students, together with (3) expected innovative enhancements of game playability, resulting in guidelines for further research.