Understanding the motivations, participation, and performance of open source software developers: A longitudinal study of the Apache projects

From AcaWiki
Jump to: navigation, search


Citation: Jeffrey A. Roberts, Il-Horn Hann, Sandra A. Slaughter (2006) Understanding the motivations, participation, and performance of open source software developers: A longitudinal study of the Apache projects. Management Science (RSS)

doi: 10.1287/mnsc.1060.0554


Tagged: Apache (RSS), open source (RSS), motivations (RSS), FLOSS (RSS)


Summary:

Understanding the motivations, participation, and performance of open source software developers: A longitudinal study of the Apache projects was published in the 2006 special edition of Management Science on FLOSS. It tacks very closely to the psychological literature on motivation -- and on intrinsic versus extrinsic motivation in particular -- to create a theory driven model of why FLOSS developer contribute. The model is evaluated using both archival data and survey responses from Apache project contributors and then fit using structural equation modeling. Perhaps most notably, the authors find no evidence of crowding out in FLOSS.

The project is framed from the perspective of a project leader or manager aiming to attract contributors and describes how understanding individual motivation may be key to understand how this plays out. The authors offer a reasonably complicated model (shown in a figure in the paper) that explains the role between use-value motivations, intrinsic motives, extrinsic motives, status and opportunity motives, performance, past performance, education, experience, and participation in the project.

The authors test their data using a combination of survey of 288 participants from Apache, Jakarta, and the Apache XML projects (30% response rate) and additional information on the number submissions and lines of code added to the Apache project over time.

The authors offer a long series of explicit hypotheses:

  • H1A: Contributors' intrinsic motivation to participate in FLOSS are negatively related to being paid to participate (i.e., simple crowding out). (No support)
  • H1B: Contributors' intrinsic motivation to participate in FLOSS are negatively related to use-value motivations to participate. (No support)
  • H1C: Contributors' intrinsic motivation to participate in FLOSS are negatively related to their status motivations motivations to participate. (No support)
  • H2A: Being paid to participate in FLOSS projects is positively related to contributors' use-value motivations to participate. (No support)
  • H2B: Being paid to participate in FLOSS projects is positively related to contributors' status motivations to participate. (Support)
  • H3: Being paid to participate in FLOSS projects is positively related to contributors' level of participation. (Support)
  • H4A: Contributors' use-value motivations to participate in FLOSS projects are positively related to contributors' level of participation. ('Opposite direction)
  • H4B: Contributors' status motivations to participate in FLOSS projects are positively related to contributors' level of participation. (Support)
  • H5: Contributors' intrinsic motivation to participate in FLOSS projects are positively related to their level of participation. (Not supported)
  • H6: Contributors' intrinsic motivation to participate in FLOSS projects are positively related to their performance ranking (e.g., their place in the hierarchy of Apache). (Support)
  • H7: An increase in contributors performance ranking is associated with an increase in their intrinsic motivations to participate. (Not supported)
  • H8A: An increase in contributors performance ranking is associated with an increase in their use-value motivations to participate. (Not supported)
  • H8B: An increase in contributors performance ranking is associated with an increase in their status motivations to participate. (Supported)

The findings are that status motivations seem to enhance intrinsic motivation (and result in more participation) and, most strongly, that there is no crowding out effect. Additionally, there is relationship between participation levels and intrinsic motivation.

Theoretical and practical relevance:

The paper has been cited more than 100 times, mostly by the literature on peer production. It remains one of the articles on FLOSS that most closely and explicitly tries to build on the psychological literature in general and the literature on motivation in particular.