Week 05 – System Development
Narayan Ramasubbu, Chris F. Kemerer (2016) Technical Debt and the Reliability of Enterprise Software Systems: A Competing Risks Analysis. Management Science 62(5):1487-1510.
People know that software reliability is crucial for business, especially enterprise software systems, such as ERP, Server OS. In daily practices, however, developing and patching a software to keep reliable is costly due to the interdependencies and potential for conflict between the underlying, vendor-supplied platform and the customizations done by individual clients. Some engineers then take shortcuts to avoid huge working load. Then the reliability of systems shrinks since the shortcuts to rapidly deliver the functionality demanded by business trade off the potential longer-term benefits of appropriate software design investments. These designs are called technical debt.
Few study focused on the empirical explore technical debt since the Interdependencies, by nature, make it difficult to measure and assess the impact of technical debt on system reliability. The author proposed a classification of systems failures-client errors and vender errors-to overcome the measurement difficulty. Analyzing a longitudinal data set spanning the 10-year life cycle of a COTS-based enterprise software system deployed at 48 different client firms with a survey analysis, the authors find that 1) technical debt decreases the reliability of enterprise systems and 2) modular maintenance was approximately 53% more effective than architectural maintenance in reducing the probability of a system failure due to client errors and 3) it had the side effect of increasing the chance of a system failure due to vendor errors by approximately 83% more than did architectural maintenance activities.
The authors showed that how firms could evaluate their business risk exposure due to technical debt accumulation in their enterprise systems, and assessed the estimated net effects, both positive and negative, of a range of software maintenance practices.
The paper investigates software process diversity and what are its antecedents.It further links software process diversity and organizational process compliance to software project performance.To this end,the paper next defines key process area(KPA) in any project and the various dimensions (separation,variety,disparity) of KPAs implemented in any project.To get relevant measures of the constructs the authors use an innovative discovery based research approach where they dig deep into literature first and then see its relevance to actual practice to construct various measures (examples include code size,degree of customer involvement etc).The data comes from a leading multinational software company which was renowned in process innovation.Focus groups,meetings and individual interviews were carried out to find out variations such as plan versus process based process design.Next the paper goes on to develop 5 hypotheses connecting software process diversity,organizational compliance and fit between process diversity and process compliance to project performance.The DV for the analysis project performance consists of two variables:software productivity and quality.The model is divided into 2 parts:the first one measures the impact of antecedents on process diversity and findings lead credence to the hypotheses that requirements volatility,design novelty and customer involvement all had a positive and significant effect on process diversity.From the second analysis the authors conclude that indeed better fit between process diversity and process compliance efforts yielded higher productivity and quality respectively.On the other hand control variables show that requirement volatility,software code and large team size have a negative effect on performance.Similarly,too much customer customer involvement also leads to a negative effect.The paper does a bunch of robustness checks to confirm the results.Thus the paper draws on prior work on organizational and demographic diversity and extends the literature by looking at software diversity from lens of variety and disparity dimensions.It extends the argument on planning versus agility and informs the literature to look beyond these.Implementable guidelines for practitioners are also highlighted.
Paper: Langer, N., Slaughter, S.A. and Mukhopadhyay, T., 2014. Project managers’ practical intelligence and project performance in software offshore outsourcing: A field study. Information Systems Research, 25(2), pp.364-384.
Practice intelligence is the ability to resolve project related work problems that are unexpected and can not be resolved by established processes and frameworks. This study examines the role of project manager’s practical intelligence in the performance of software offshore outsourcing projects. The authors posit that project managers could apply practical intelligence to effectively address and resolve unexpected incidents, and therefore the level of practical intelligence positively affects project performance. The authors further theorize that project complexity and familiarity contribute to its information constraints and the likelihood of critical incidents in a project, thereby moderating the relationship between practical intelligence and project performance.
The authors explore the research question by analyzing a detailed longitudinal data that record project and personnel level activities on 530 projects completed by 209 project managers. The authors employ the critical incidents methodology to assess the practical intelligence of the project managers who led these projects. The authors consider several factors that influence task complexity and requires different level of practical intelligence such as Technological Complexity (Software size and schedule compression), Organization Complexity (team size and team dispersion), and Task familiarity. The findings indicate that project managers’ practical intelligence has a significant and positive impact on project performance (cost performance and client satisfaction). Further, projects with higher complexity or lower familiarity benefit even more from project managers’ practical intelligence.
Kang, K., Hahn, J., & De, P. (2017). Learning Effects of Domain, Technology, and Customer Knowledge in Information Systems Development: An Empirical Study. Information Systems Research, 28(4), 797-811.
Though there have been a myriad of developments and improvements to programming languages, development methods and tools, and formal training in information systems and technologies, information systems developments (ISDs) are still plagued by performance concerns including failing to adhere to schedules and timetables as well as budgets. Recent research has proposed and explored how capitalizing on learning (or experience) curves could improve ISDs through the theory of transfer of learning. Effectively, ISD projects are difficult due to their disparate tasks, teams, and complexity that vary across projects. To the extent that these projects capitalize on existing knowledge or experience amongst team members, they will be better able to meet performance deadlines and expectations. Using archival data outlining 497 ISD projects that included 2,393 unique employees at a prominent global IT services company from 2005 to 2007, Kang, Hahn, and De (2017) were able to explore how ISD project teams are able to learn and transfer knowledge from prior to new projects and under which conditions learning effects are stronger or weaker.
The theory behind Kang et al.’s (2017) study highlights how employees are able to accrue learning effects through repeated experience of identical elements across tasks (i.e., the environmental perspective). Results indicated that ISD project teams’ experience from prior projects benefits performance (operationalized as development effort and scheduled delay) on subsequent projects when there is an overlap in knowledge or experience. Moreover, ISD project teams can have domain (worked in the domain before), technology (worked with same technology), and customer experience (having dealt with the customers before). These three experience areas are substitutive and become either stronger or weaker depending on the projects’ team and task complexities. For example, the technology experiences a project team may have are less beneficial (i.e., becomes weaker) when the project has high technological complexity.
Kang, K., Hahn, J., & De, P. (2017). Learning Effects of Domain, Technology, and Customer Knowledge in Information Systems Development : An Empirical Study. Information Systems Research, 28(4), 797–811
Investigation of ISD performance by identifying the factors associated with its increase (or decrease) has recently been the focus of much attention in the information systems literature. Some literature proposes that ISD performance can be improved by taking advantage of learning curve (or experience curve) effects. This study examines how ISD project teams learn and under what conditions the learning effects are stronger or weaker. The authors conceptualize ISD as the application of various elements of domain, technology, and customer knowledge, and propose that the units of the same experience should be these knowledge elements.
The research questions are: (1) Do ISD project teams benefit from prior experiences in domain, technology, and customers in increasing project performance? (2) whether and how experiences in the three types of knowledge relate to one another. (3) How does ISD project complexity moderate the learning effects of these three types of experience?
From a dataset collected from a prominent global IT services company, this study finds that ISD project teams’ experience in prior projects translates into performance gains for the current ISD project when the prior and current projects share the same domain, technology, or customer knowledge elements. In addition, learning effects of domain, technology, and customer knowledge are substitutive for one another and that these learning effects become stronger or weaker depending on the extent of ISD projects’ team and task complexities.
The study makes significant contributions to the ISD literature on learning effects and the roles of domain, technology, customer knowledge, and project complexity, as well as to the general organizational learning literature. It also provides important managerial insights into practical concerns such as project staffing and knowledge acquisition for ISD organizations.
Project managers’ ability always affects the performance of software project directly, especially in the case of offshored software projects which spans locational and organizational boundaries, engendering greater information gaps and higher uncertainties, practical intelligence (PI) of PMs can be significantly helpful in solving difficulties and challenges.
The paper considers projects from an information processing perspective. Specifically, offshored software projects are prone to gaps between information processing needs and capabilities that pose severe management challenges, often lead to unexpected impediments to project success. In order to mitigate the negative effects of incomplete information, projects can use approaches like standard project management techniques, select appropriate PMs and team members.
Then the authors adopt the taxonomy of PI for IT professionals to that for PMs. They use four metrics to evaluate it, tasks, career, self, and others. Then, they consider that PM’PI may benefit projects differently based on the projects’ features. One of them is project complexity which consists of technological complexity and organizational complexity. A more complex project impose more severe information processing challenges, therefore requiring a higher PI for PM. Specifically, higher project complexity would increase the information needs of a project; In contrast, higher familiarity with the task and between stakeholders would increase the information capabilities in a project. Project complexity consists of technological complexity and organizational complexity.
The authors conduct a field study at a leading software outsourcing vendor in India. Then, they use the critical incidents approach to assess PM’S PI and collect data on more than 500 software outsourcing projects, those data including the complexity, familiarity, outcomes of each project and a series of control variables. OLS regression models are established to test hypotheses, results support most of the hypotheses.
This paper has several contributions. It demonstrates the project performance effects of PI and in the certain context of offshore outsourcing, it also conceptualizes and quantifying PM’s PI, and examine its efficacy in different contexts.It also has practical implications in choosing and training PM.
Dr. Narayan Ramasubbu at University of Pittsburgh will be joining us as a guest speaker on Feb 14. His website is at https://sites.google.com/site/narayanramasubbu/. Please read his papers and come up with a question to him.
|Langer et al. (2014)||Leting, Jack|
|Ramasubbu et al. (2015)||Sid|
|Ramasubbu and Kemerer (2016)||Joe|
|Kang et al. (2017)||Xi, Heather|
We will discuss Subramanyam et al. (2012) when we have time. Please be sure to read it as well.