Predictors of creative computing participation and profiles of experience in two Silicon Valley middle schools

Abstract

Examination of the “digital divide” has increasingly gone beyond the study of differences in physical access to computers to focus on individuals’ use of technological tools for empowered and generative uses. In this research study, we investigated the relationship between access to tools and experience with creative production activities. Our participants included 160 8th grade learners from two public middle schools. The local communities represented by the two schools differed in parent education levels, proportion of recent immigrants, and average family income. Findings indicated substantial variability in students’ history of creative production experiences within both communities. Three sets of analyses were completed. First, the two school populations were compared with respect to average levels of student creative production experience, access to tools at home, use of learning resources, frequency of technology use, and access to computing outside of their home. Second, correlates of variability in individuals’ breadth of experience with creative production activities were explored across both schools through a regression analysis. The resulting model indicated that students’ experience was best predicted by the number of technology tools available at home, number of learning resources used, frequency of computer use at home, and non-home access network size. In a third analysis, profiles of experience were created based on both breadth and depth of experience; the resulting four groups of students were compared. More experienced students utilized a broader range of learning resources, had access to more tools at home, taught a wider range of people, and were more confident in their computing skills. The groups did not differ in their self-reports of interest in learning more about technology.

Introduction

Design-oriented activities are believed to play a special role in learning to adapt computing tools for one’s own purposes (Ber, 2006, Kafai, 1995, Papert, 1980). Projects that involve the production of a personally meaningful artifact can offer the motivation that drives persistence and the setting of learning goals, as designers work to create what they imagine (Barron, 2006). Within such projects learners frequently encounter implementation challenges that provide opportunities to develop the knowledge, skills, and intellectual capabilities that underlie what has been called technological fluency (National Research Council, 1999). Web design, game making, robotics, programming, animation, and movie making are all examples of projects that are likely to be fluency-building and that youth find compelling (Barron, 2006, Resnick et al., 1998, Walter, et al., 2007).

It is not clear that opportunities for these types of creative production activities are equally distributed across more and less affluent communities, raising concern that the benefits of computing resources will accrue for those who are already most economically advantaged. Individual differences in computing experience are one manifestation of what has been called the digital divide (Hargittai, 2003). Initially the term “digital divide” was defined with respect to computer ownership or basic access to the Internet. More recent definitions have reflected a multidimensional construct capturing inequities in how people use computing tools and how skilled they are (DiMaggio et al., 2004, Hargittai, 2008). Use and skill can vary as a function of income, age, ethnicity, gender, education level, or geographic location, as these variables frequently reflect differences in access to tools and learning resources (Barron, 2004, Hargittai and Hinnant, 2008, Warschauer, 2000).

In this paper we report on a study designed to contribute data on the relationship between experience with creative production activities and access to tools and resources. Our research is guided by a learning ecology framework grounded in the assumption that learning takes place across the life spaces of home, school, and community (Barron, 2004) and that the proper unit of analysis is the individual learner and the multiple settings in which the learner spends time. The learning ecology framework draws on ecological perspectives from psychology as well as constructs developed from sociocultural and activity theory. Ecological perspectives emerged from a desire to better articulate the interdependencies between variables at the child and environment levels, and acknowledge the tight intertwining of person and context in producing developmental change (Bronfenbrenner, 1979, Cole, 1996, Lerner, 1991, Lewin, 1951, Rogoff, 2003). Sociocultural, activity, and situative learning theorists (Cole, 1996, Cole, 2000, Engeström, 1987, Greeno, 1989, Lave and Wenger, 1991, Pea, 1993, Rogoff, 2003, Vygotsky, 1978) foreground the role of tools that have been created by prior generations as critical mediators of cognitive and social practices such as language, writing, and other representational systems.

Ecological metaphors have recently been applied to other studies of technologically rich environments (Brown, 2000, Looi, 1999). For example, Nardi and O’Day (1999) introduced the idea of an information ecology. In their view, “an information ecology is a system of people, practices, technologies and values in a local environment. Like their biological counterparts, information ecologies are diverse, continually evolving, and complex.” The current learning ecology definition shares with the definition of information ecologies the idea that both relational and material resources are important in any socio-technical ecology (Brown, 2000, Nardi and O’Day, 1999) and it implies a dynamic learning system open to multiple influences. In line with this perspective, it is key to understand learning as made possible by configurations of resources in the forms of social relationships, tools, informational resources, and activities. Consequently for this study we designed metrics and analyses that would provide information about an individual’s access to computing tools and learning resources at home, in school, in the community or on the Internet. In addition to examining correlations between experience and access to resources, we investigated variables such as confidence, interest, and knowledge sharing, which are likely influenced by experience with creative production activities.

Below we summarize some of the empirical literature that documents correlations between a number of demographic variables and measures of use, access, and skill.

Despite narrowing gaps in access to computing tools at home and at school as a function of demographic variables, there is growing concern that new information technologies may contribute to further inequalities along economic, cultural, or gender lines because of differential use, attitudes, or skill (DiMaggio et al., 2004, Hargittai, 2008). This concern is fueled by recent studies that suggest that differential use of computing tools is related to demographic variables, including gender and socioeconomic status (SES). In a socioeconomically and ethnically diverse college freshman sample, positive correlations between parent education levels and creative contributions were found and males were more likely to post original artwork online (Hargittai & Walejko, 2008). Similarly, analysis of the 2003 US census data showed that students whose parents have any graduate education or whose family income is $75,000 or greater are approximately twice as likely as their peers to use the Internet to complete school assignments and find information (DeBell & Chapman, 2006).

Differences in types of technology use associated with demographic variables are mediated by access to learning opportunities, both those provided in formal settings such as schools or after-school clubs, and informally in peer networks or home settings. Both quantitative survey data and qualitative comparative studies have found that while students with low-SES use school computers more frequently than do their high-SES counterparts in math and English courses (which often use drills), high-SES students are the main users of technology in science courses where computers are often used for more sophisticated tasks, such as simulation and research (Becker, 2000, Margolis et al., 2008, Warschauer, 2000). Additionally, previous work has found that even when children have similar levels of home access to computers, those from higher socioeconomic backgrounds are more likely to experience educational gains from the resource than children from lower SES backgrounds (Livingstone & Helsper, 2007). Other comparative case studies have shown that lower SES schools had more problems keeping computers working and spent less money on professional development that would help teachers learn to incorporate technology with instruction in meaningful ways (Warschauer, 2000).

Studies of technology-related course offerings have shown differential rates of high level computing electives between low-SES and high-SES schools. While courses offering computer science or programming are offered in 10–14% of schools in the top three SES-based quartiles, these courses are taught in only 5% of the bottom SES-based quartile (Becker, Ravitz, & Wong, 1999). Goode, Estrella, and Margolis (2006) noted that in Los Angeles, the College Board certified Advanced Placement computer science course was disproportionately offered in schools serving more affluent districts. Even within groups that have high access to technology, experience with creative production activities is linked to use of learning resources such as courses in or out of school, reading material, or tutorials (Barron, 2004).

Family practices also influence interest and learning. Studies of middle class families have shown that the amount of parent–child coactivity around computing predicts a child’s interest and engagement in computing (Simpkins, Davis-Kean, & Eccles, 2005). Similarly, a qualitative study of parent roles in learning in an affluent sample identified a variety of ways that parents nurtured creative production projects through collaboration, lending or buying learning resources such as books, by brokering opportunities such as summer camps, and by directly explaining concepts to their child (Barron, Martin, Takeuchi, & Fithian, 2009).

Finally, design experiments, a form of research that first develops innovative environments and then studies how learning occurs within them (Brown, 1992), provide evidence that divides can be bridged through intentional interventions. Projects that create opportunities for youth in community-based computing clubs or schools show that even when home access is low, resources in the forms of tools and mentors can lead to increased engagement in programming, music creation, graphic design, and other creative production activities (Barron, 2006, Barron et al., 2006, Peppler and Kafai, 2007, Resnick et al., 1998). Together, the studies reviewed in this section support the claim that although mean differences as a function of demographic groups are frequently observed, there is also variability within demographic groups that needs to be better understood.

Research done by Hank Becker suggests that technological divides along economic lines are worsened by “community effects”. He argues that because of residential segregation by SES, children living in low-SES families tend to live in neighborhoods where they are less likely than children living in more affluent communities to have access to a computer through a neighbor or friend (Becker, 2000). This line of reasoning can be extended to consider the importance of social learning networks (National Research Council & Institute of Medicine, 2000). Access to peers and adults who have experience, interest, and expertise is likely to be as important or more important than access to tools. Given the documented spread of technical knowledge in co-located communities, it is important to consider the nature of school or neighborhood communities in terms of how widely expertise is distributed. Peer groups at school are often the source of inspiration for the development of production-oriented hobbies such as web design, animation production, or robotics (Barron, 2006, Chandler-Olcott and Mahar, 2003). The likelihood of finding a classmate with such interests will be related to the proportion of experienced peers in a school. At this point, we know of no studies that have provided data comparatively examining distributions of experience profiles with creative production activities across schools. While co-located communities are important to understand, from a learning ecology perspective it is important to consider learning opportunities as distributed across life settings (Barron, 2004, Barron, 2006). Accordingly, in this research we examined access to learning resources that may be found at home, school, in the community, or on the Internet.