1
Learning through active participation and the effect of leadership in
OSS projects
Jorrit Blom 2030055
University of Groningen, Faculty of Economics & Business Duisenberg Building, Nettelbosje 2
9747 AE Groningen, The Netherlands Supervisor:
dr. J.Q. Dong Second Supervisor:
P.M.M. de Faria
2 Abstract
Interest in Open Source Software ( OSS) has risen considerably in the last decade with increased interest from both academics and corporations. With developers working on voluntary basis and governance being based on a meritocracy, motivation and leadership differ significantly from
traditional work environments. This paper uses three separate aspects of leaders within OSS projects to determine how leadership can affect self- improvement of contributors as one their motivations to participate in OSS. Data was collected from a dataset involving 181 OSS projects found on
SourceForge and used for statistical analysis. It was found that increasing the number of leaders on projects and leaders that have long durations of being part of OSS have positive effects on learning through active participation.
3
Introduction
4 articles focusing on this topic. Similarly, Hars and Ou (2002) found within their research that 70.9% of their respondents chose “improving my programming skills” when asked why they participate in open source projects. Many companies and also universities have started to see the added value of OSS and how it can improve someone’s skill level (Sen, 2007). Several schools already started pilot studies for classes for software engineers to capitalize on the potential for learning opportunities at low costs (Sowe & Stamelos, 2007).Moreover, it is argued that programmers can develop skills through active participation and interacting with more experienced and knowledgeable contributors by working on OSS projects (Hemetsberger & Reinhardt, 2006).
5
How can different aspects of leadership and governance within OSS projects affect learning through active participation in OSS?
Additional research into what motivates and what effects the motivation of volunteers to contribute is important since one of the biggest problems for OSS projects is reduced participation resulting in over 75 percent of OSS projects disappearing over time (Fang & Neufeld, 2009). To answer this question this paper adopts the following structure. First of all, a theoretical background will be discussed to address the academic literature on motivation within OSS, self-improvement by participation and governance and leadership within OSS projects. After the theoretical background the development of the hypotheses will be discussed which are developed based on academic literature. The next parts will discuss the methodology, measures and the results of the statistical analysis. Finally, the paper will conclude with a discussion and conclusion of the results and the implications and limitations of the paper.
Theoretical background
OSS motivations
In the field of psychological academic literature the theory of self- determination identifies and makes a distinction between two different parts of individuals motivation, which are intrinsic and extrinsic motivation (Ryan & Deci, 2000). Intrinsic motivation is defined as the drive to do an activity in order to experience pleasure and satisfaction, either enjoyment-based which is the drive to obtain satisfaction through participating in an activity, or obligation-based which is the drive to meet the morals, values. and ethics dictated by an individual (Li, Tan & Teo, 2012).Whereas, extrinsic
motivation is defined by a drive to take action to attain rewards, including career, prestige and
6 found in the OSS literature. One of the most identified motivations for participation within literature on OSS are career opportunities by demonstrating your skills and building a reputation (Lerner & Tirole, 2002). Corporations have noticed the potential of OSS and can use it to scout programmers with potential that they might be able to recruit (Sen, 2007). Voluntary participants in OSS can use the work they do on projects as a testament to their skills. Reputation can be very important to the programmers working on OSS projects since there is a culture of meritocracy within the field of OSS (Blincoe et al., 2016). Therefore, programmers can gain status and recognition by working on a project and inputting contributions in order to gain importance and authority, making them more motivated to keep working on open source projects (Lerner & Tirole, 2002). Another commonly identified motivation within the literature on open source is the desire for self-improvement among contributing members (Hars & Ou, 2002; Martinez-Torres & Diaz-Fernandez, 2013). Besides just being able to learn through participation it also allows for the possibility to help design the technical side of the project and be able to experiment with new technology or software (Von Krogh,
7 motivation is when projects are being commercialized for private gain with unfair and unequal appropriation (Benkler, 2002).Programmers take a certain pride in their work and do not want to see that work be commercialized by others. One of the main reasons for the start of open source software in the 1970s was the fact that software companies tried to include publicly available software into their proprietary software (Qureshi & Fang, 2010).
Self improvement through active participation
Many developers contributing to OSS projects do this on a voluntarily basis for several possible reasons, one of which being the opportunity to improve their own skills while working on projects. In community of practice literature actively experimenting and participating is shown to increase the learning of skills and tacit knowledge (Hemetsberger & Reinhardt, 2006). Improving programming abilities involves a high degree of actively working on tasks like debugging, maintaining and extension of programs and does not just stop at passive learning from reading the written codes of the programs (Ye & Kishida 2003). OSS projects are therefore an good opportunity to improve skills on both higher and more basic levels since OSS projects involve both more difficult tasks and basic tasks (Au, Carpenter, Chen & Clark, 2009). Huntley (2003) considers OSS projects as a good example of adaptive learning, since members continuously have to adapt and respond to changes in the state of the project and software while over time responding becomes easier and faster.
8
Tullio & Staples, 2013). Another method for learning through participation is debugging the software of the project. The process involves several tasks, for example modifying program codes, reviewing and repeatedly scanning, all of which can be considered lower-level learning tasks that can enhance skills (Au, Carpenter, Chen & Clark, 2009). By actively participating in these tasks it gives new members the opportunities to get in touch with experienced programmers and ask questions in order to learn from their expertise and to learn how to use, modify and extend programs and systems (Ye & Kishida, 2003). Universities and schools have also recognized this potential as a low cost teaching and learning tool and have started integrating pilot studies in OSS in their curriculum for software engineers (Sowe & Stamelos, 2007).
Governance and leadership in OSS.
Open source software governance can be defined by using the definition from Markus (2007) who defines OSS governance as ‘the means of achieving the direction, control and coordination of wholly
or partially autonomous individuals and organizations on behalf of an OSS development project to which they jointly contribute’’ (Markus, 2007). Governance and leadership within open source
software projects are different from more conventional teams and organisations. Open source software communities and the governance it involves are almost always based on virtual teams that can be distributed worldwide geographically. Combined with members contributing and
9 bottom layer and based on merits it is possible to move up the metaphorical pyramid. At the top of the governance structure are the leaders of the projects and sometimes skilled people referred to as elders (Blincoe et al., 2016). Leaders have the right to coordinate the projects meaning they decide on what gets added, when it gets added and which tasks have to be done and by who (Li, Tan & Teo, 2012). Elders are common on longer running projects and are contributors with long tenures that may have previously been in charge of projects but now adopt an advisory role at the top (Blincoe et al., 2016). The second structure of governance can be described as a series of concentric circles with the centre consisting of the core developers that can be compared to leaders in charge of coordination. Around the centre of core developers there is a ring that includes the maintainers that hold
responsibility over a small part of the project, for example a module. The final ring involves the contributors doing smaller but necessary tasks like bug fixing, reporting, handling the support but also includes the general users (Blincoe et al., 2016). De Laat (2007) found two different types of governance structures within open source software projects. The first model is the democratic model that can be characterized by lower degrees of centralization and coordination through democratic discussions in order to divide the tasks together (de Laat, 2007). The second model is the autocratic model which is the opposite of the democratic model. Autocratic governance models are
characterized by higher degrees of centralization in which self appointed leaders, most commonly the creators of the project, assign the responsibilities. Which means the autocratic model is more of a top-down governance model, whereas the democratic model is bottom-up (de Laat, 2007). However, in both cases reputation and merits generally have high influence in the coordination decisions. The downside of the autocratic model shows when the self-appointed leaders do not judge based on accomplishments and when hierarchy is to strict, all of which can lead to abandonment of the project by volunteers and ultimately failure (Bezroukov, 1999). Leadership in OSS projects means having to manage volunteers and keep them motivated for sustained participation. Two varieties of
10 & Bhattacherjee, 2012). When using a responsibility management style the leaders themselves distribute the various responsibilities as they see fit. Whereas, participation management has a deeper focus on management through analysing all the contributions and distributing responsibilities based on participation and contributions (Midha & Bhattacherjee, 2012).
Hypotheses development
The governance and leadership in OSS projects differs from regular teams and organizations but also varies greatly between different projects. According to path-goal theory the behaviour of leaders and the work environment they create is instrumental in affecting motivation of their subordinates and how tasks are performed (Li, Tan & Teo, 2012). Lerner and Tirole (2002) found four tasks that leaders perform in order for the project to run smooth. First of all, leaders create the vision that is established by the software codes that are examples of the leaders expertise (Lerner & Tirole, 2002). Secondly, they are in charge of attracting new programmers and creating a challenging environment where programmers have the opportunities to work on challenging problems and have the
opportunities to improve the software code (Lerner & Tirole, 2002). Thirdly, leaders are in charge of dividing projects into modules which allows contributors to work independently from others (Lerner & Tirole, 2002). Finally, they are responsible for resolving and avoiding conflicts which can lead to people stopping to contribute to the project. Therefore, leaders have an important impact on
coordination and the community surrounding the OSS projects. A leader, in the OSS community, coordinates the project which mean they can decide when additions are released, allocate tasks and even if necessary reward or punish individuals (Yan, 2014). One of the main motivations for contributors found in academic literature was the chance to improve your own skills through
11 motivations of contributors. This paper examines three different aspects of leaders in OSS projects that might affect learning through active participation.
Size of leadership
Most OSS projects have an structure based on meritocracy meaning the most experienced and the most skilled generally have more important roles within the projects governance (Capra, Francalanci & Merlo, 2008). Therefore, leaders of the projects generally have more experience and skills and are at the core of development compared to most of the other developers (Blincoe et al.,2016). In most cases the developers that have contributed most also have higher influence within the project and on the tasks that have to be accomplished. Active participation of new members creates opportunities for them to interact with other more knowledgeable and skilled developers, and gives them
legitimate access to the expertise therein (Lakhani & von Hippel, 2003). Furthermore, learning from peers was found to be the most effective in all different learning states (Singh, Yong & Youn, 2011) The existence of OSS communities enables new participants to inquire about a variety of aspects of the OSS systems and to acquire help in using, understanding, modifying, and extending the systems (Lakhani & von Hippel, 2003). Since leaders are selected on merits and reputation they usually have higher skill levels than the average contributor. OSS projects are an ideal opportunity for
programmers wanting to improve to get in touch with and ask questions to higher skilled
programmers, in this case leaders of OSS projects (Capra, Francalanci & Merlo, 2008). More project leaders means more skilled programmers that can increase self-improvement of lower skilled
programmers wanting to improve. Furthermore, higher numbers of leaders also means
12 Hypothesis 1
The size of leadership within the OSS project has a positive effect on learning through active participation
OSS experience of leadership
Project leaders in OSS are generally speaking programmers that have built a reputation for themselves. Since participation is on a voluntary basis members can change projects or work on multiple projects quite easily. Meaning that many programmers contributing to open source software work on many projects and do not necessarily stick to one (Lakhani and von Hippel, 2003). The benefit of this high mobility within OSS projects is that it allows programmers to experience many different projects and gain experience in different fields (Li, Tan & Teo, 2012). Furthermore, it also allows individuals to create more social contacts with other programmers of which they can learn. A member with more direct contacts is more likely to be connected with other powerful actors in the network, potentially receiving information of higher quantity and quality than individuals with less direct contacts( Li, Tan & Teo, 2012). This means working on a variety of projects can increase the skill and knowledge base of programmers and allow them to create larger social networks and
13 Di Benedetto, 2013). In addition, a team leader's experience, as measured by counting the number of prior projects a team leader has led, has been found to be a key driver of improvement project success (Easton and Rosenzweig, 2012). Overall, working on different projects can increase the experiences and skills of leaders that they can use to coordinate and assist other members. Hypothesis 2
Projects with more experienced leaders that on average have joined more projects has an positive effect on active participation in tasks and therefore learning through participation
Duration of leadership in OSS
Not all programmers like to work on multiple different projects to gain experience but prefer to work on projects they are committed to. Programmers do not only gain experience by working on multiple projects but of course also by just working on OSS in general for an extended period of time whether that is a singular or multiple projects. One of the main drivers of gaining authority within OSS projects is your reputation. And reputations are built over longer durations of time in which you can showcase and improve your skills. It was found that people with longer tenures in OSS are perceived to be more knowledgeable and more likely to become a leader within projects (Bitzer & Geishecker, 2010). The longer the duration you spend working on OSS projects the larger your skill base,
14 developers spend on OSS the larger their knowledge repositories become and the easier it is to recognize data sets and algorithms (Singh, Yong & Youn, 2011).
The longer leaders are active in the OSS communities the more status and recognition they earn increasing their effectiveness as leaders which can positively influence contributors feelings towards the project (Xu, Jones & Shao 2009). Since meetings and contact are only on a virtual basis, it is vital that there exist skilled arbiters in the form of leaders whose ability and authority are above question in order to inspire and resolve disputes (Fitzgerald, 2006). Therefore, the more experience there is present within the project the more other developers can learn from each other. In this case that means experiences and skills that were gained by just being a part of OSS for a longer duration of time. Leaders that have been active on OSS for longer durations will be able to share knowledge and lessons they learned with contributors that are willing to learn.
Hypothesis 3
The duration of leadership working on OSS has an positive effect on active learning through participation
Conceptual model
15 Figure 1. Conceptual model
Methodology
Data collection
16 which is the dependent variable. To identify which observations were useable a dummy variable was used to see whether the project had included the dependent variable or not. After excluding all observations that did not include the dependent variable 181 observations were left that were all used in the statistical analysis
Measures
Dependant variable
The dependant variable that was used as a measure for learning through active participation is the variable closed tracker tickets. OSS projects commonly use issue tracking systems to keep track and coordinate requests, support, handle defect fixing and all sorts of tasks involved with managing a software project. Tracker tickets are the requests and tasks within the project that involve all kind of usually smaller tasks to improve the project. This might involve support for other users, debugging or making patches. Overall, these are tasks that in most cases are handled by less experienced
contributors due to the easier nature of the tasks meaning it is an opportunity to learn by contributing (Au, Carpenter, Chen & Clark, 2009). It was found that projects need a high level of efficiency when it comes to defect removal and enhancement of functionality in order to become a high performing project (Ghapanchi & Aurum, 2012).Tracker tickets all represent active tasks that have to be done by contributors on the project in order to manage and enhance the project. This variable measures how many tracker tickets were actually closed by people working on the project. These kind of tasks are perfect for people wanting to improve their skills and familiarize themselves with the community and programs. The name of the variable used is Sum_Closetracker which is the sum of all tracker tickets that have been closed on the project.
Independent variables
17 variable num_leaders which measures how many leaders are present within the project. OSS
leadership experience is measured by the variable leader_avgProjectJoined. It measure how many projects are joined on average by the leaders of the observed project. The more projects they have joined the higher their experience is. Finally, duration of leadership is measured by the variable leader_avgDaysExp which looks at how many days on average the leaders of an observed project have been active on SourceForge regardless of how many projects they were involved with.
Control variables
While this study focuses on aspects of leadership and their effect on the motivation of learning, other variables might also have an influence on motivations of contributors. Therefore, this paper controls for several factors that might affect the motivation of self-improvement.
Number of features: One reason programmers might contribute to OSS projects is to fulfil their own needs (Bogers, Afuah & Bastian, 2010). Specific features might motivate members to contribute and keep contributing. Furthermore, a larger number of features on a project can also mean there are more tasks to perform possibly affecting the dependent variable. Meaning more opportunities to improve your own skills working on these features and more tracker tickets that need to be closed.
Number of categories: It is possible that programmers only want to improve their skills based on certain categories of projects. Furthermore, It was found that the number and type of categories applicable to the project can have significant effects on the choice to participate and keep
participating in the available tasks within the project (Subramaniam, Sen & Nelson, 2009). If there are more categories that the projects has to appeal to, there might be more tracker requests.
18 2013). This variable is measured by the number of licenses, the greater the number of licenses the more restrictive the project.
Number of languages: Literature has shown that besides regular programming skills, some people also use the OSS projects to practice languages like English or Spanish (Hars and Ou, 2002). Furthermore, it was found that the number of languages that projects are translated into can have a positive effect on human resource attraction which in this case means developers (Ghapanchi & Tavana, 2015). Therefore, the more different languages are available the more interesting a project can be to use for improvement of your language skills.
Number of programming languages: The main goal of this paper is to research the effects on learning through active participation. One of the main skills that can be improved through
participation in OSS is related to programming skills which means you have to learn how to utilize different programming languages (Hars and Ou, 2002). Furthermore, it was found that the number of programming languages that are available in projects can have a positive effect on human resource attraction which in this case means developers (Ghapanchi & Tavana, 2015). Therefore, the more programming languages are available in the project the more attractive it is for self-improvement through active participation.
19 Number of interfaces: The main goal of this paper is to research the effects on learning through active participation. Besides learning to use certain programs contributors looking for self
-improvement also use it to familiarise themselves with programs and interfaces. It was found that the number and type of interfaces used in the project can have significant effects on the choice to
participate and keep participating in the available tasks within the project (Subramaniam, Sen & Nelson, 2009). Therefore, the more interfaces are available in the project the more attractive it is for self-improvement through active participation.
Size of project: File size indicates how large the projects is. Larger projects have more features included, more codes to analyse and more tasks to perform. The larger the project grows the more challenges it opens up for the project to coordinate the necessary tasks(Wang, Shih, Wu & Carroll, 2015).Therefore, it may affect the dependent variable, the sum of closed tracker tickets, since larger file size mean more task of maintaining the software quality by for instance debugging.
Age of project: De Laat (2007) found different evolutionary stages that OSS projects might go through. The older an project is the further they are in their evolution and complexity when it comes to coordination can change drastically( De Laat, 2007). Furthermore, it was found that the higher the age of a project the more interest from users and developers increases (Sen, Singh & Borle, 2012). More interest in the project means more requests for features and possibilities for enhancement through bug fixing. Therefore, the age of an project might have an effect on closing tracker tickets, since older projects are more complex when it comes to coordination and the availability of tasks.
Sum of downloads: When the scale of projects grows larger it opens them up to new challenges in order to coordinate tasks and maintain the quality of the software (Wang, Shih, Wu & Carroll, 2015).
The higher the sum of downloads the larger the scale and interest in the project. This larger interest leads to challenges for coordination and more requests for tasks to maintain the quality like
20 defects because there are more users to notice possible bugs (Ghapanchi, Wohlin & Aurum,
2014).Therefore the sum of downloads can affect the number of tracker tickets and the number of tasks that have to be accomplished within the project.
Sum of open tracker tickets: Since the dependent variable measures how many of the open tracker tickets in this variable have actually been closed, there might be a relationship between these two variables. Furthermore, the more tracker tickets are open the more open tasks there are to perform on the project.
Results
Descriptives
The first step that was taken was to run a test to show the descriptive statistics of the dataset that was used. The descriptive results are shown in figure 2. As can be seen in figure 2, the dataset included 15 variables with 181 different projects observed. Something noticeable is the standard deviations of the different variables and that there are several variables with very high standard deviations. This means that the data contains a lot of variation which is certainly the case since the variance in certain variables should differ a lot. Variables like the sum of downloads and the project size in kilobytes will obviously differ substantially per observed project and large differences between these projects are to be expected. After examining the results further several extreme values or outliers were identified and since removing the outliers would only create new ones, it was decided not to delete them. Based on the large standard deviations in certain variables outliers are expected and removing them would only result in loss of data.
Model summary
21 R squared of the model is 0.662 and the adjusted R squared is 0.634 meaning that over 60 % of the total variance of the data is explained within the model. Even though there is no determined
threshold for an acceptable R squared, an R squared off over 60 percent can be seen as fairly high and therefore satisfactory (Taylor, 1990). A possible explanation of the high R squared is the fact that many control variable have been included that might explain for the variance. In order to check whether it is truly the independent variables that have the main effect on the dependent variables, the same test was done again only this time without the three independent variables. Without the three independent variables the R squared is 0.205 and the adjusted R squared is 0.154. It can be
22 Correlations
In order to get a better look at the relationships between the different variables in the model a Pearson correlation test was performed. The results of the test can be found in figure 3. The
24 Regression results
25 The three main independent variables are Num_Leader, leader_avgDaysExp and
leader_avgProjectJoined. Hypothesis 1 used the variable Num_Leader to test whether more leaders would have an positive effect on the dependent variable. The significance score of the number of leaders was found to be 0.00 meaning hypothesis 1 is supported with a confidence level of 0.05. Furthermore, the standardized coefficient Beta indeed shows that this variable has a positive effect with a standardized Beta of 0.723. Hypothesis 2 focuses on the number of different projects leaders have joined on average and was proposed to have an positive effect on the dependent variable. Hypothesis 2 has a significance score of 0.907 meaning that there is no support for the relation that was argued in hypothesis 2. Hypothesis 2 is rejected. Hypothesis 3 looks at the total time leaders have spent working on OSS projects and is proposed to have a positive effect on the dependent variable. The significance score of the number of leaders was found to be 0.063 meaning hypothesis 1 is supported with a confidence level of 0.1. This means the hypothesis finds moderate support since it uses a higher confidence level of 0.1. Furthermore, the standardized coefficient Beta indeed shows that this variable has a positive effect with a standardized Beta of 0.152. Besides the three
26
Variable significant at the 0.05 level* Variable significant at the 0.1 level**
Discussion and conclusion
27 literature on OSS identifies the motivation to learn and improve skills as one of the most important reasons for people to contribute in OSS(Lerner & Tirole, 2002; Hars & Ou 2002). By working on OSS project you can improve certain skills or familiarize yourself with the programs and other programmers by actively participating. The leaders in OSS projects coordinate the projects and decide what gets added and what tasks have to be done meaning they have significant influence in what every member can do(Li, Tan & Teo, 2012). This paper looked at three different aspects of leadership within OSS projects and how they affected the dependent variable measuring learning through active participation. Hypothesis 1 focused on the size of leadership of the OSS projects and stated that more leaders can positively influence learning effects from working on OSS projects. The reasoning behind this is that leaders within OSS are generally chosen based on reputation and merits in the project or other projects (Blincoe et al.,2016). Which means they tend to have higher skill levels and greater understanding than contributors still trying to improve their skills. More leaders therefore means a larger base of skilled programmer from which to learn and ask specific questions in order to improve (Lakhani & von Hippel, 2003). Hypothesis 1 was confirmed by the statistical results. Hypothesis 2 focused on the experience of leaders and the beneficial effects to motivation of contributors. Moreover, it focuses on experience by working on multiple different projects and posits that experience of leaders in different projects has a positive effect on learning by contributors. The reasoning behind this is that working on multiple different projects will give the leaders a wider base of knowledge due to different kinds of experience gained from different projects (Easton and
Rosenzweig, 2012). However, hypothesis 2 was not found to be statistically significant and was therefore rejected. A possible explanation might be that leaders who worked on many projects do not necessarily have higher skill profiles. They have experience in multiple projects and balanced
28 their level of programming skills, making them a jack-of-all-trades instead of a specialist which might be less interesting for individuals looking for specific specialists.Hypothesis 3 focused on the duration of leaders within OSS projects and looked at the general amount of time spend working on OSS. The more time one spends working on something the more they learn by actively participating and doing the tasks. Therefore, leaders that have spent a longer duration of time working on OSS can have higher skill levels and understanding about how everything works and therefore higher ability to guide others in the project increasing the positive effect on learning of other contributors.
Hypothesis 3 was supported at a higher confidence level of 0.1.
Managerial implications
OSS has been steadily growing over the years with even increasing attention from businesses using it in their own operations. Therefore, this paper has several implications for managers from both
normal businesses and people involved on voluntary basis in OSS. First of all, this paper contributes to the academic literature focusing on motivations that drive OSS participation. Which is important since most participation is on voluntary basis and many projects disappear due to lack of
contributors. Therefore, increased understanding of what motivates and what effects the motivation is important to keep contributors from leaving the projects. This papers main purpose was to examine how different aspects of leadership within OSS projects effects the learning experience and
29 motivation to learn since there are more leaders with abilities to learn from meaning increasing the amount of skilled leaders in OSS projects can be a good managerial decision. Secondly, it was found that the duration of OSS leadership or the amount of time leaders had spent working on OSS in general is important. Therefore, managers should keep in mind the total time contributors have spent working on OSS when selecting leaders.
Limitations and future research
In this section the main limitations of this paper are discussed and suggestions for future research are provided. Starting with the first and perhaps most important limitation, the
30 Acknowledgements
31 References
Au, Y. A., Carpenter, D., Chen, X., & Clark, J. G. (2009). Virtual organizational learning in open source software development projects. Information & Management, 46(1), 9-15.
Benkler, Y. (2002). Coase's Penguin, or, Linux and" The Nature of the Firm".Yale Law Journal, 369-446.
Bezroukov, N. (1999). Open source software development as a special type of academic research: Critique of vulgar Raymondism. First Monday, 4(10).
Blincoe, K., Sheoran, J., Goggins, S., Petakovic, E., & Damian, D. (2016). Understanding the popular users: Following, affiliation influence and leadership on GitHub. Information and Software
Technology, 70, 30-39.
Bitzer, J., & Geishecker, I. (2010). Who contributes voluntarily to OSS? An investigation among German IT employees. Research policy, 39(1), 165-172.
Bogers, M., Afuah, A., & Bastian, B. (2010). Users as Innovators: A Review, Critique, and Future Research Directions. Journal of Management, 36(4), 857-875.
Capra, E., Francalanci, C., & Merlo, F. (2008). An empirical study on the relationship between software design quality, development effort and governance in open source projects. Software
32 De Laat, P. B. (2007). Governance of open source software: state of the art. Journal of Management
& Governance, 11(2), 165-177.
Di Tullio, D., & Staples, D. S. (2013). The governance and control of open source software projects. Journal of Management Information Systems, 30(3), 49-80.
Dong, J. Q., Wu, W., & Zhang, S. Y. (2015). Impact of innovation speed on user interest in open source software, Working Paper, Hong Kong University of Science and Technology, Hong Kong.
Easton, G. S., & Rosenzweig, E. D. (2015). Team leader experience in improvement teams:: A social networks perspective. Journal of Operations Management, 37, 13-30.
Fang, Y., & Neufeld, D. (2009). Understanding sustained participation in open source software projects. Journal of Management Information Systems, 25(4), 9-50.
Fitzgerald, B. (2006). The transformation of open source software. Mis Quarterly, 587-598.
Ghapanchi, A. H., & Aurum, A. (2012). The impact of project capabilities on project performance: Case of open source software projects. International Journal of Project Management, 30(4), 407-417.
33 Ghapanchi, A. H., Wohlin, C., & Aurum, A. (2014). Resources contributing to gaining competitive advantage for open source software projects: An application of resource-based theory. International
Journal of Project Management, 32(1), 139-152.
Gini, A., & Green, R. M. (2014). Three Critical Characteristics of Leadership: Character, Stewardship, Experience. Business and Society Review, 119(4), 435-446.
Giuri, P., Rullani, F., & Torrisi, S. (2008). Explaining leadership in virtual teams: The case of open source software. Information Economics and Policy,20(4), 305-315.
Giuri, P., Ploner, M., Rullani, F., & Torrisi, S. (2004). Skills and openness of OSS projects:
Implications for performance. Paper provided by Laboratory of Economics and Management (LEM),
Sant'Anna School of Advanced Studies, Pisa, Italy in its series LEM Papers Series with number 2004, 19.
Hars, A., & Ou, S. (2002). Working for Free? Motivations for Participating in Open-Source Projects.
International Journal of Electronic Commerce, 6(3), 25-39.
Hemetsberger, A., & Reinhardt, C. (2006). Learning and knowledge-building in open-source communities a social-experiential approach. Management learning, 37(2), 187-214.
34 Lakhani, K. R., & Von Hippel, E. (2003). How open source software works:“free” user-to-user assistance. Research policy, 32(6), 923-943.
Lerner, J., & Tirole, J. (2002). Some Simple Economics of Open Source.Journal of Industrial
Economics, 197-234.
Li, Y., Tan, C. H., & Teo, H. H. (2012). Leadership characteristics and developers’ motivation in open source software development. Information & Management, 49(5), 257-267.
Markus, M. L. (2007). The governance of free/open source software projects: monolithic, multidimensional, or configurational?. Journal of Management & Governance, 11(2), 151-163.
Martínez-Torres, M. R., & Diaz-Fernandez, M. C. (2014). Current issues and research trends on open-source software communities. Technology Analysis & Strategic Management, 26(1), 55-68.
Midha, V., & Bhattacherjee, A. (2012). Governance practices and software maintenance: A study of open source projects. Decision Support Systems,54(1), 23-32.
Peng, G., Mu, J., & Di Benedetto, C. A. (2013). Learning and Open Source Software License Choice. Decision Sciences, 44(4), 619-643.
35 Raja, U., & Tretter, M. J. (2012). Defining and evaluating a measure of open source project
survivability. Software Engineering, IEEE Transactions on,38(1), 163-174.
Roselli, D. S., Lorch, J. R., & Anderson, T. E. (2000). A Comparison of File System Workloads. In USENIX annual technical conference, general track(pp. 41-54).
Ryan, R. M., & Deci, E. L. (2000). Intrinsic and extrinsic motivations: Classic definitions and new directions. Contemporary educational psychology, 25(1), 54-67.
Sen, R. (2007). A strategic analysis of competition between open source and proprietary software. Journal of Management Information Systems, 24(1), 233-257.
Sen, R., Singh, S. S., & Borle, S. (2012). Open source software success: Measures and analysis. Decision Support Systems, 52(2), 364-372.
Sine, W. D., Mitsuhashi, H., & Kirsch, D. A. (2006). Revisiting Burns and Stalker: Formal structure and new venture performance in emerging economic sectors. Academy of Management
Journal, 49(1), 121-132.
Singh, P. V., Yong, T., & Youn, N. (2011). A Hidden Markov Model of Developer Learning Dynamics in Open Source Software Projects. Information Systems Research, 22(4), 790-807
Sowe, S. K., & Stamelos, L. G. (2007). Involving Software Engineering Students in Open Source Software Projects: Experiences from a Pilot Study. Journal Of Information Systems Education,
36 Subramaniam, C., Sen, R., & Nelson, M. L. (2009). Determinants of open source software project success: A longitudinal study. Decision Support Systems, 46(2), 576-585.
Taylor, R. (1990). Interpretation of the correlation coefficient: a basic review.Journal of diagnostic
medical sonography, 6(1), 35-39.
Von Hippel, E. (2005). Democratizing innovation: The evolving phenomenon of user innovation. Journal für Betriebswirtschaft, 55(1), 63-78.
Von Krogh, G., Haefliger, S., Spaeth, S., & Wallin, M. W. (2012). Carrots and rainbows: Motivation and social practice in open source software development.MIS quarterly, 36(2), 649-676.
Wang, J., Shih, P. C., Wu, Y., & Carroll, J. M. (2015). Comparative case studies of open source software peer review practices. Information & Software Technology, 671-12
Xu, B., Jones, D. R., & Shao, B. (2009). Volunteers’ involvement in online community based software development. Information & management, 46(3), 151-158.
Yan, L. (2014). Social capital characteristics of open source software opinion leaders. Journal of
Computer Information Systems,54(4).
