Modeling Learners in Educational Games: Relationship Between Playing and Learning Styles

Background: Video games are very well known for their intrinsic adaptivity; as they adapt the gameplay to the player’s preferences and rhythm. In addition to adapting the gameplay, educational games should also adapt the learning content to match the learner’s competencies, preferences, playing and learning styles. Determining the playing style in an educational game is made possible by collecting certain metrics and information susceptible of monitoring the player’s interactions in the game. However, it is still a challenge to assess the learner’s learning style. Aim: This study examines the correlation between learning and playing styles. It has been acknowledged that both playing and learning styles are related to personality. After examining the personality traits of each style of both Kolb’s learning styles and Bartle’s playing styles, it was hypothesized that there would be a correlation between the two. In that purpose, a quantitative research was conducted to explore the relationship between the two taxonomies. Method: One hundred high school students majoring in science in Morocco have completed Kolb’s learning style inventory (version 3.1) as well as the Bartle test questionnaire developed by Andreasen and Downey. The statistical correlation between the two taxonomies was investigated using cross-tabulation, Chi-square/Fisher’s exact test and Pearson coefficient. Results: Results revealed a relevant interdependency between Kolb’s learning styles and Bartle’s playing styles. The ‘convergent’ type was found to correlate with Bartle’s explorer type; the majority of assimilators adopt the killer type in a game; accomodators tend to be achievers and divergers prefer to be socializers . In terms of learning styles, it was noticed that the majority of participants adopted the converging and the assimilating learning styles; which is in line with what it is indicated in Kolb’s inventory. Results have also shown that learning and playing styles are gender independent. It was concluded that the learning style can be predicted, based on the playing style.