Understanding contributor behaviour within Free/Libre/Open Source Software communities: A socialization perspective
Attracting a large number of new contributors has been seen as a way to ensure the survival, long-term success, and sustainability of Free/Libre/Open Source Software (FLOSS) communities. However, this appears to be a necessary, but not a sufficient condition, as the well-being of FLOSS communities also relies on members performing behaviours that nurture and protect the community. Despite a large body of research on FLOSS communities, few studies have been undertaken to explore the influence of a participant’s socialization experience on their contribution behaviour. In addition, there has been relatively little research that has adopted a community-level view of FLOSS community participants’ contribution that goes beyond the mere notion of writing lines of code. The purpose of this study is to develop and rigorously test a socialization model that predicts contributor behaviour in the FLOSS community context. Drawing upon theories of socialization and citizenship behaviours from organizational behaviour research, this research develops and tests two separate but related research models. The first model proposes the direct impact of socialization factors on two performance-related dependent variables: task performance and community citizenship behaviours. The second model hypothesizes a mediating effect of two proximal socialization variables, social identification and social integration, between the socialization factors and the dependent variables. An exploratory study involving eleven FLOSS community leaders, managers, and experienced members was first conducted, to investigate the key variables that characterize FLOSS community newcomer socialization experience as well as the various instances of citizenship behaviours that are specific to the FLOSS community context. The analysis of the interview data revealed the existence of six socialization variables: task segregation, task purposefulness, interaction intensity, mentoring, joining structuredness, and supportiveness. Two sets of FLOSS community citizenship behaviours (CCB) were identified drawing on the citizenship behaviour literature. The first set, labelled CCB-I, comprised citizenship behaviours directed towards the benefit of individuals. The second set, CCB-P, included citizenship behaviours directed towards the benefit of the project. The findings were integrated in the two conceptual models. Subsequently, a research instrument was developed, following an extensive purification process that consisted of card sorting and expert review rounds, and a survey pretest. A pilot study assessed responses from 46 FLOSS contributors from two large FLOSS communities. Overall, the scales demonstrated high reliability and showed adequate construct validity. The analysis of the pilot study suggested the existence of a third CCB dimension, named CCB-C, that characterizes citizenship behaviours that are oriented towards the benefit of a project’s community. The main study was based on an online survey involving 327 respondents from twelve large FLOSS communities. Using Partial Least Squares (PLS), the collected data was used to test the two models. The results showed the overall superior predictive capability of the model hypothesizing the mediating effect of both social identification and social integration. Task performance was found to be directly predicted by task purposefulness as well as by interaction intensity and supportiveness through the mediation of social identification. Meanwhile, CCB was found to be impacted by the direct effect of task segregation and task purposefulness, and by interaction intensity and supportiveness through the mediation of both social identification and social integration. The existence of the third CCB dimension, CCB-C, was confirmed.