Pair Programming for Middle School Students: Does Friendship Influence Academic Outcomes? Linda Werner University of California Santa Cruz, CA 831-459-1017 linda@soe.ucsc.edu Jill Denner, Shannon Campe, Eloy Ortiz ETR Associates 4 Carbonero Way Scotts Valley, CA 831-438-4060 {jilld, shannonc,eloyo}@etr.org Dawn DeLay, Amy C. Hartl, Brett Laursen Florida Atlantic University Boca Raton, FL ABSTRACT Research shows the benefits of pair programming for retention and performance in computing, but little is known about how relationship dynamics influence outcomes. We describe results from our study of middle school students programming games using Alice and pair programming. From our analysis using statistical procedures that take into account the interdependence of pair data, we found evidence for partner influence moderated by the role of confidence over improvements in Alice programming knowledge in friend partnerships but not non-friend partnerships. We discuss implications for researchers and educators. Categories and Subject Descriptors K.3.2 [Computers and Education]: Computer and Information Science Education – Computer science education. Keywords Pair Programming, Game Programming, Middle School, Alice, Friendship, Computational Thinking. 1. INTRODUCTION Educators and researchers are trying to change the ‘face of computing’. Those at the K-12 level are hopeful that initial programming environments such as Alice and Scratch [22] and techniques such as collaboration via pair programming and game programming will make computing accessible to all students at an early age. Since computer science (CS) is not yet widely accepted as one of the K-12 core topics [3], we need to provide evidence- based results that show using these programming environments and techniques holds promise for introducing CS concepts and bringing real gains in computational thinking (CT) to K-12 students. We recently presented results supporting the introduction of CS concepts using Alice and game programming for middle school youth [24]. In this paper, we describe results pertaining to the use of friendship in the partnerships, referred to as dyads, with respect to pair programming and gains in aspects of CT. As a review, “CT is a problem-solving process that includes (but is not limited to) the following characteristics:  Formulating problems in a way that enables us to use a computer and other tools to help solve them.  Logically organizing and analyzing data,  Representing data through abstractions such as models and simulations,  Automating solutions through algorithmic thinking (a series of ordered steps)  Identifying, analyzing, and implementing possible solutions with the goal of achieving the most efficient and effective combination of steps and resources,  Generalizing and transferring this problem solving process to a wide variety of problems.” [11] We ran semester-long game-programming courses called iGame that involved 189 middle school students using Alice and pair programming across four semesters (in and after school). iGame is a study where we address many research questions including the following: Under what conditions does pair programming benefit the learning outcomes of middle school students programming games using Alice? Are the dyadic data interdependent? We looked closely at pair programming to determine if there were moderating effects of this form of dyadic learning processes. To address this research question, we conducted surveys containing questions about student confidence in computing and Alice programming knowledge. Additionally students rated the degree to which their partner was a friend. In this paper, we first discuss prior work with pair programming, followed by a description of the iGame course including participant demographics, procedures of the study, measures, and the analysis process. Next, we list the results of the analyses using statistical procedures that take into account the interdependence of dyadic data and discuss these results. The paper concludes with a discussion of future work. 2. PRIOR WORK A long history of research describes the benefits of collaborative learning [4] and specifically pair programming for individual performance and persistence rates at the college level [6][18][26] but there are very few studies of pair programming in K-12 [15][25]. Building on the research of Vygotsky [23], socio Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. SIGCSE’13, March 6–9, 2013, Denver, Colorado, USA. Copyright 2013 ACM 978-1-4503-1868-6/13/03 …$15.00.