Article Preview
TopIntroduction
In current technology-oriented education two trends, among others, can be discerned: an increasing attention for game-based learning (GBL, see Wouters, van Nimwegen, van Oostendorp & van der Spek, 2013) and personalised learning (Beetham & Sharpe, 2013; Tseng, Chu, Hwang & Tsai, 2008). With respect to GBL reviews indicate that GBL is not always as effective and motivating as is always assumed (Wouters et al., 2013) or that these outcomes can only be realised when certain instructional conditions are met such as the inclusion of opportunities for players to explicate the acquired knowledge (Wouters et al., 2013; see also Clarke, Tanner-Smith, & Killingsworth, 2016). In addition, Erhel and Jamet (2013) have suggested that the contradictory findings regarding GBL can be explained by the different methodologies used in the studies, the variety of topics and learning situations and individual learner characteristics. The latter factor connects with personalized learning which can be defined as instruction in which the pace and the instructional approach adapt to the educational needs of the learner (Thomas, 2016).
In this study GBL refers to the use of computer games for the purpose of learning, training and education. The assumption is that two processes are involved: a direct cognitive process (i.e. the selection and organisation of new information and integrate these with prior knowledge) and an indirect process whereby the motivational appeal of the game is used to expose the player longer and more engaged to the learning content in the game (see Wouters & van Oostendorp, 2017). Computer games and therefore also GBL environments have a number of features that make them suitable for applying personalized learning. To start with they are highly interactive: players act in the game world and receive feedback as the consequences of their actions are reflected in the game world. Second, the digital nature makes it possible to record, unobtrusively and unlimited, performance data that can be used to adjust the game world accordingly. For example, in a GBL environment the complexity of a learning task could be decreased real-time when the performance data indicate that the previous tasks were too difficult. In this context the results of stealth assessment are promising (Shute, Ke, & Wang, 2017).
However, for personalized learning not only real-time adaptivity based on performance data is important but also a better understanding with regard to student characteristics such as prior knowledge, preference for an instruction method, interests, learning style etc. Some of these characteristics can be adapted by learners themselves, for example, selecting a context in which the learning material is presented, the composition of an avatar or even the preference for an instruction method. In a study using a game-based environment with arithmetical problems, children who received a personalized version (e.g., they could choose a context for the arithmetical problems and the names of characters) outperformed children in the non-personalized version in motivation as well as learning (Cordova & Lepper, 1996, see also Wouters & van Oostendorp, 2017). Other learner characteristics such as prior knowledge and learning style are more complex and must be determined in an objective and valid manner.