[Paper Review] Planning Gamification Strategies based on User Characteristics and DM: A Gender-based Case Study

Armando M. Toda, Wilk Oliveira+4 more|arXiv (Cornell University)|May 22, 2019

Educational Games and Gamification24 references33 citations

TL;DR

The paper uses a large-scale survey and association rule mining to derive gender-specific gamification element recommendations for education, highlighting Objective-Progress relationships and gender-differentiated element preferences.

ABSTRACT

Gamification frameworks can aid in gamification planning for education. Most frameworks, however, do not provide ways to select, relate or recommend how to use game elements, to gamify a certain educational task. Instead, most provide a "one-size-fits-all" approach covering all learners, without considering different user characteristics, such as gender. Therefore, this work aims to adopt a data-driven approach to provide a set of game element recommendations, based on user preferences, that could be used by teachers and instructors to gamify learning activities. We analysed data from a novel survey of 733 people (male=569 and female=164), collecting information about user preferences regarding game elements. Our results suggest that the most important rules were based on four (out of nineteen) types of game elements: Objectives, Levels, Progress and Choice. From the perspective of user gender, for the female sample, the most interesting rule associated Objectives with Progress, Badges and Information (confidence=0.97), whilst the most interesting rule for the male sample associated also Objectives with Progress, Renovation and Choice (confidence=0.94). These rules and our descriptive analysis provides recommendations on how game elements can be used in educational scenarios.

Motivation & Objective

• Motivate the development of data-driven gamification design tailored to gender differences in education.
• Identify which game elements are most preferred overall and by gender.
• Discover relationships among game elements and how they relate to user gender.
• Provide actionable recommendations for educators on selecting and combining gamification elements.

Proposed method

• Collect a large survey (N=808 raw; 733 valid) on preferences for 19 gamification elements used in education.
• Assess data normality and compare genders with non-parametric tests (Mann-Whitney).
• Compute association rules using the Apriori algorithm to uncover element relationships by gender (min support 10% for males, 20% for females; min confidence 90%).
• Validate rules using additional metrics: conviction, lift, and leverage.

Experimental results

Research questions

• RQ1 How do gender differences affect preferences for gamification elements in education?
• RQ2 What are the strongest associations among gamification elements for male and female learners?
• RQ3 Can data-driven association rules inform gender-specific gamification design guidelines for teachers?
• RQ4 Which core element (e.g., Objectives) serves as a central hub in gender-based rules?

Key findings

• Females show strongest associations linking Progress, Acknowledgement, and Data with Objectives (confidence 0.97, lift 1.63).
• Males exhibit prominent rules where Progress appears with Level, Renovation, and Choice, but fewer rules reach 90% confidence for social elements.
• Across both genders, the Objective element is central in rules, indicating its broad acceptance in educational gamification design.
• Descriptive analysis reveals females favor Renovation, Puzzles, and Sensation, while males favor Cooperation, Competition, and Classification.
• Association rules indicate Objective–Progress as a common, well-accepted pairing for both genders.