Scenarios in a serious game for training communication skills

Abstract

Communication skills are designated as one the keys skills for the 21st century. Practice by training on the job or in role play with a simulated character are effective ways to learn communication skills. In serious games to practice communication skills, a player performs a dialogue with a virtual character to learn and practice a communication skill in a particular situation, for example ‘Breaking bad news’.

In my dissertation, I consider three aspects of a serious game, namely: develop generic components and tools; analyse learning aspects and game mechanics; and evaluate and improve pedagogical aspects. I investigate these aspects in the context of Communicate, a serious game for practicing communication skills.

We successfully created a generic dialogue integration component that can be used to develop and integrate dialogue scenarios seamlessly in a serious game. We also successfully created a smells tool that a scenario author can use to detect potential scoring problems in a scenario.

We used information within a scenario to create a scenario specific corpus and use this corpus to match open text input from a player during gameplay. We call our method scenario specific corpus method (SSCM). We evaluated SSCM against generic Natural language processing (NLP) methods and found that SSCM slightly outperformed generic NLP methods for matched text but underperformed for unmatched text.

We compared student feedback to various interventions in blended teaching. Our results suggest that in designing and employing serious games, it is crucial to investigate pedagogical aspects such as peer teaching.