DataRPG: Improving student motivation in data science through gaming elements

The amount of data is exploding, and by all estimations will continue its exponential growth. This has led to an increased demand for engineers and data scientists. However, courses imparting crucial ideas from data design, administration and maintenance to data analytics and knowledge discovery remain under-subscribed due to perceived difficulty, low interest and motivation. This motivation can be represented by three constructs: interest in the topic, personal control and difficulty of the exercise. Role-playing games (RPGs) have been proven to stimulate students interest in analysis and problem solving when faced with varying tasks. This paper investigates the integration of specific game-based elements and pop-culture references into course delivery and assessment to boost interest and learner motivation in an elective postgraduate database systems course at The University of the West Indies. The experimental setup for assessment of students' progress in this course uses an experience point (XP) level-based system adapted from popular RPGs. To further stimulate interest, pop-culture references to specific animated series' and fantasy game genres are used. A student can choose to take one of multiple paths for their learning, representing different tools/platforms that can be used in implementation and analysis. Students can attempt more than one path for more experience and higher levels, although it is not expected. Completion of each path indicates formative “mastery” of the selected data science tool/platform. Summative assessments uses a “raid-party system” where students who have been encouraged along different paths are now asked to join together to 1. analyse given datasets and capture their findings in the format of an academic paper, and 2. build a complete database application for a particular scenario. Formative feedback is done in groups to leverage scaffolding. Measurement of the students' activity and interest is done using automatically generated activity logs from the online delivery course system as well as administration of an online electronic questionnaire. Both of these instruments are used to determine how effective this approach is in influencing students' interest and motivation. More than 5 times the historical enrollment numbers have been observed and initial results point to improvement in students' attitudes toward the content. This action research would be used to inform content delivery and further investigations would determine the impact on student performance.