Figure 1: Students using MultiLearn Exploring Collaborative Learning Methods for Multiplayer Educational Games in Low Resource Environments Heather Underwood 1 , Sunil Garg 1 , Clint Tseng 1 , Leah Findlater 2 , Richard Anderson 1 , Joyojeet Pal 2 { 1 Computer Science & Engineering, 2 The Information School } University of Washington { hmu2@cs, skgarg@cs, cxlt@cs, leahkf@u, anderson@cs, joyojeet@u } .washington.edu ABSTRACT To facilitate learning in low-resource environments where children must share a single computer, we developed MultiLearn, an educational, single-display groupware system. Prior work has suggested that computer games can be effective for fostering learning [5] and qualitatively supports the benefit of collaboration over competition in a multiple-input, shared display environment [2]. Through MultiLearn we have begun to explore the benefits of introducing collaboration into an innately competitive setting (computer games) for learning environments in low-resource areas. We present preliminary findings from a study of 192 students in Bangalore, India, exploring the impact of differing degrees of collaboration and competition within MultiLearn. Author Keywords Single display groupware, shared computing, multiple input, education, collaboration, competition, developing regions ACM Classification Keywords H.5.2 Information interfaces and presentation (e.g. HCI): User Interfaces. General Terms Experimentation, Design, Human Factors, Performance INTRODUCTION Computer use in low-resource learning environments typically consists of a many-to-one scenario, in which multiple children share a single computer. Sharing a keyboard and mouse often results in dominance patterns and unequal access to educational material [1]. To address this issue, researchers have introduced systems to allow for multiple users to use the same computer through multiple mice [1] building on past work on single display groupware [3]. Taking this work further, we present MultiLearn (Figure 1), an educational classroom computer game using multiple 10-key numeric keypads and a split-screen design. The MultiLearn software consists of several games that drill students on math, spelling, and other subjects. It uses adaptive questioning based on individual student performance and includes a teacher platform for curriculum development. The durable, portable, low-cost hardware provides a feasible solution for developing regions. Prior research suggests that both collaboration and competition impact children's engagement and learning gain, and that ways of combining the impacts of both are an important step ahead in the multiple input learning space [2]. Consequently, a critical design component in MultiLearn is the incorporation of collaboration in the interactions between individual users. We have designed several game configurations that enforce individual versus team scoring and individual versus team input. To understand the impact the different modes have on students' performance and collaboration outcomes, we conducted a study with 192 grade school students in Bangalore, India. We present here preliminary findings comparing a purely competitive mode and a combination mode of collaboration and competition. Our findings illustrate that a mix of competition and collaboration results in improved performance and evidence of students working together to solve a problem. In examining important mechanics in the actual collaborative process, we find that issues such as the need to challenge stronger children, domination by aggressive children, and gender issues in mixed groups further support the fundamental premise of this research—that an appropriate balance between competition and collaboration is likely to offer the most in learning gains in multiple input scenarios. STUDY METHODOLOGY We conducted a study to compare the impact of different collaborative setups, varying individual versus team play, as well as individual versus shared keypads. The goal was to explore how to effectively introduce collaboration into a competitive educational game in the context of multiple input devices. In this paper, we focus on two game modes that best illustrate the difference in performance between a strictly competitive mode and a mode that physically promotes collaboration: No Collaboration (M4) Each student was given a keypad and played a purely competitive game against three other players. Physical Collaboration (M2) Pairs of students shared a keypad and completed questions together, competing against other pairs. Copyright is held by the author/owner(s). CSCW 2010, February 6–10, 2010, Savannah, Georgia, USA. ACM 978-1-60558-795-0/10/02. 475