Arun’s research in the area of Software Development and Project Management has examined how innovations in software development technologies and methodologies can address problems in software development/maintenance; how formal and informal controls can mitigate different risks in the development process; how complex software development projects that require collaboration across expertise, role, and cultural boundaries can be effectively managed; and how online communities for open source software development can be governed to sustain the community. These projects have been in collaboration with development groups, Fortune 100 firms with large IT organizations, and leading offshoring vendors and their client firms.

Selected Publications

The adoption rate of computer-aided software engineering (CASE) technology continues to be low among information systems departments (ISDs). Some ISDs have reported significant hurdles in propagating CASE usage, while documenting the advantages of the technology. We construct and empirically test a theoretical model to explain CASE adoption behavior. Factors considered include need pull (environmental instability of the ISD and performance gap of the ISD), technology push (internal experimentation and learning from external information sources), and the adoption context (top-management support for the IS function, CASE championship, training availability, and job/role rotation). A national survey of 2,700 ISDs resulted in 405 usable responses for the data analysis.Our analysis suggests a reasonable fit between the model and the data. The results indicate that the need-pull factors do not directly promote CASE adoption behavior. Performance deficit promotes CASE championship behavior while negatively affecting other elements of the adoption context. The instability of ISDs, where the very existence of the ISD may be in question, negatively affects all elements of the adoption context. Learning about CASE from external information sources directly promotes CASE adoption. Both technology push factors positively affect all four elements of the adoption context. Of the contextual elements, CASE training availability, CASE championship, and job/role rotation positively affect CASE adoption behavior. Top management support does not affect CASE adoption behavior, which suggests that such support may be more critical for postadoption stages of the diffusion process.

The availability of high-quality software is critical for the effective use of information technology in organizations. Research in software quality has focused largely on the technical aspects of quality improvement, while limited attention has been paid to the organizational and sociobehavioral aspects of quality management. This study represents one effort at addressing this void in the information systems literature. The quality and systems development literatures are synthesized to develop eleven quality management constructs and two quality performance constructs. Scales for these constructs are empirically validated using data collected from a national survey of IS organizations. A LISREL framework is used to test the reliability and validity of the thirteen constructs. The results provide support for the reliability and validity of the constructs. A cluster analysis of the data was conducted to examine patterns of association between quality management practices and quality performance. The results suggest that higher levels of institutionalization of all quality management practices are associated with higher levels of quality performance. Our results also suggest that key factors that differentiated high-and low-quality performing IS units include senior management leadership, mechanisms to promote learning and the management infrastructure of the IS unit. Future research efforts directed at causally interrelating the quality management practices should lead to the development of a theory of quality management in systems development.

We identify top management leadership, a sophisticated management infrastructure, process management efficacy, and stakeholder participation as important elements of a quality-oriented organizational system for software development. A model interrelating these constructs and quality performance is proposed. Data collected through a national survey of IS executives in Fortune 1000 companies and government agencies was used to test the model using a Partial Least Squares analysis methodology. Our results suggest that software quality goals are best attained when top management creates a management infrastructure that promotes improvements in process design and encourages stakeholders to evolve the design of the development processes. Our results also suggest that all elements of the organizational system need to be developed in order to attain quality goals and that piecemeal adoption of select quality management practices are unlikely to be effective. Implications of this research for IS theory and practice are discussed.

Software projects can often spiral out of control to become "runaway systems" that far exceed original budget and schedule projections. The behavior that underlies many runaway systems can best be characterized as "escalation of commitment to a failing course of action." The objectives of this study were to: (1) understand the extent to which IS projects are prone to escalate, (2) compare the outcomes of projects that escalate with those that do not, and (3) test whether constructs associated with different theories of escalation can be used to discriminate between projects that escalate and those that do not. A survey was administered to IS audit and control professionals and, to establish a baseline for comparison, the survey was designed to gather data on projects that did not escalate as well as those that did escalate. The results of our research suggest that between 30% and 40% of all IS projects exhibit some degree of escalation. Projects that escalated had project outcomes that were significantly worse in terms of perceived implementation performance and perceived budget/schedule performance, as compared to projects that did not escalate. Using constructs from theories that have been used to explain the escalation phenomenon, we were able to test various logistic regression models for their ability to discriminate between projects that escalate and those that do not. To construct our models, we explored constructs derived from self-justification theory, prospect theory, agency theory, and approach avoidance theory. While constructs derived from all four theories were significant in logistic regression models, the completion effect construct derived from approach avoidance theory provided the best classification of projects, correctly classifying over 70% of both escalated and non-escalated projects.

Improving the capability of the development process has emerged as an important strategy for addressing recurring problems in software development, such as poor quality, high development costs, and long delivery lead times. While a number of normative software process models have been proposed, limited theory development has occurred. This study adopts a knowledge management perspective to examine the antecedents of software process capability. Synthesizing prior research in total quality management (TQM), software engineering and knowledge management, it is posited that creation of process knowledge and embedding the knowledge in the process are important antecedents of software process capability. Two models that relate these antecedents to software process capability are tested using data collected from information systems (IS) units in Fortune 1000 firms. The first model is a direct-impact model, which posits that both knowledge creation and knowledge embedding have a direct and positive impact on software process capability. The second model is a mediated-impact model, which posits that knowledge embedding completely mediates the relationships between knowledge creation and software process capability. The results provide strong support to the mediated-impact model suggesting that the extent to which knowledge is embedded in the process has a significant positive effect on software process capability. The results also suggest that the stock of knowledge embedded in the process is influenced by the extent to which knowledge is created in the context of the development process. Implications of these findings for IS research, theory, and practice are discussed.

Systems development processes have received significant negative publicity due to failed projects, often at large costs, and performance issues that continue to plague IS managers. This study complements existing systems development research by proposing a knowledge management perspective for managing tacit and explicit knowledge in the systems development process. Specifically, it proposes that collaborative exchange and integration of explicit knowledge across phases of the development process positively influence the performance of systems development. It also suggests that process formalization not only directly impacts development performance but also moderates the performance effects of the knowledge integration factors. Data for the empirical study were collected from 60 organizations that are part of a user group or one of the world’s largest software development tool vendors.

Empirical results provide strong evidence of the importance of supporting tacit and explicit knowledge processes in systems development as well as process formalization. The findings suggest that: (i) collaborative exchange among IS employees that integrates their tacit knowledge positively impacts development performance, (ii) explicit knowledge integration in development artifacts across different phases of the systems development process positively impacts development performance, (iii) formalization of processes that establishes routines and discipline yields performance gains, and (iv) the performance effects of both collaborative exchange and explicit knowledge integration are moderated by the formalization of the process. These results have implications for how both tacit and explicit knowledge integration can be managed during systems development, and how formalization of processes complements their relationship with development performance.

Although organizations have applied a variety of practices and tools to address information systems development (ISD) performance problems, most of these approaches have focused on controlling and improving predictability of the development process. There is growing recognition that ISD is a knowledge-intensive process that requires the integration of specialized stakeholder knowledge. We develop the perspective that integration of this specialized knowledge across knowledge boundaries in the ISD process drives ISD performance. We theorize that formal and informal organizational integrative practices influence ISD performance, because they facilitate the development of boundary objects that effectively span knowledge boundaries. Results from a field study of 110 firms provide considerable support for the proposed model. This paper makes three novel contributions to the technology management literature. First, it demonstrates that knowledge integration across knowledge boundaries through boundary objects improves ISD performance. Second, it shows how formal and informal organizational integrative practices enhance the integration of specialized knowledge within and across organizational subunits. Third, it shows that the positive influence of formal and informal organizational integrative practices on ISD performance is partially mediated by knowledge integration. For engineering and technology managers, the results highlight the centrality of knowledge integration for the management of technology development. Collectively, these findings offer a novel knowledge integration-based perspective that complements prior research on systems development and new product development

Agency theory has served as a key basis for identifying drivers of offshore information system project success. Consequently, the role of relational factors in driving project success has been overlooked in this literature. In this paper, we address this gap by integrating the social embeddedness perspective and the culture literature to theorize how and why relational factors affect the success of offshore IS projects that are strategic in nature. We identify organizational and interpersonal cultural differences as critical success factors in this context. Using data from a longitudinal field study of 155 offshore IS projects managed by 22 project leaders, we found evidence of a relationship between hypothesized relational factors and two measures of offshore IS project success--namely, project cost overruns and client satisfaction-- over and above the effects of project characteristics and agency factors. Specifically, we found that information exchange, joint problem solving, and trust reduce project cost overruns and improve client satisfaction. We also found a relationship between cultural differences at the organizational and team level, and offshore IS project success. The model explained 40 percent and 41 percent of the variance in project cost overruns and client satisfaction, respectively, for projects with a client representative. For projects with no client representative, the model explained 35 percent and 37 percent of the variance in project cost overruns and client satisfaction, respectively. Collectively, the results have important theoretical and practical implications for how client-- vendor relationships should be managed when partnering with offshore firms and designing offshore IS project teams.

We integrate control theory and the information systems (IS) project management literature using a multilevel lens to theorize the cross-level effects of technical IS project risk on individual developer outcomes—performance and psychological stress—and the mechanisms by which IS project managers’ project-related knowledge attenuates this relationship. We argue that IS project managers with project-related knowledge mitigate technical IS project risk by facilitating the enactment of internal and external process controls in their IS projects. Our empirical study involves data collected from 1,230 individual developers embedded in 130 IS project teams that are managed by 20 IS project managers. Our results provide strong support for the three-level model and its set of (a) cross-level main effects of technical IS project risk on individual developer outcomes; (b) cross-level main effects of IS project manager project-related knowledge on enacted internal and external process controls; and (c) cross-level moderation of the relationship between technical IS project risk and individual developer outcomes by IS project manager project-related knowledge through internal and external process controls. Our study provides insights on how IS project management, IS project process controls, and technical IS project risk must be managed as a system of multilevel dependencies to achieve the desired developer outcomes.