Custom Class Information Systems Development Project Performance Model Based On The Risk Of A Two-stage Study

With information technology playing an increasing role in the society and economy, software developer and user have grown more heavily dependent on the successful delivery of information systems (IS). However, in overseas and domestic cases, failures of software development project are common, especially for the information systems development (ISD) projects. Most projects can not be delivered within budget and on time, the systems delivered can not meet the requirements of users. Therefore, How to manage an ISD project effectively has become an important practical problem in academic research and practice.According to researchers, a key question is how to deal with the uncertainties of software development,which has become the one of core subject in the field of IS project management.This dissertation aims at customized ISD projects from the perspective of a system provider, i.e. a software developer. Considering the nature of multiple sources for uncertainties and the dynamic natures for risk in the IS development and implementation process, this study analyzes the relationship between project inherent uncertainty, key management practices and different dimensions of project performance,and developed a two-phase risk based performance model of ISD projects. Past research pointed out that the key influential factors maybe different in different project phases. Thus, through introducing two-phase risk, the variation of risk and influential factors in the information systems development process can be dynamically analyzed and the corresponding management strategies can be discussed,which have important implications for practice of IS project management.Based on the dataset (n=181) collected from customized ISD projects of software houses in Hangzhou City ,this study carries out an empirical analysis on the research model using structural equation modeling(SEM) techniques. The sample was further grouped to study the interactions between influential factors.In addition, the research compares the ISD successful criterion from the point of view of software developers and users based on a small sample from customers.Two-phase risk refers to performance risk in development phase and performance risk in implementation phase, which respectively denotes the difficulties of estimating project outcomes during the different phases. Through an empirical test on the research model, it is found that performance risk in development phase and performance risk in implementation phase are two important intervening variables,which mediate the effects of project inherent uncertainty and four management factors(project planning and control,internal integration,user participation and top management support) on different dimentions of project performance. The findings suggest that management strategies,such as reinforcing project planning and control and internal integration, gaining the commitment and involvement of top management and key users of cutomer organization, can reduce the performance risk and improve project performance in the presence of project inherent uncertainty. In addition, the key influential management factors are found to be different in different project phases. For performance risk in development phase, key influential factors are project planning and control, internal integration and user participation. For performance risk in implementation phase, key influential factors are user participation and top management support.This study groups the total sample according to the levels of project inherent uncertainty to analyze the interactions between project inherent uncertainty and management factors. The results show that: the effect of internal integration on process performance is stronger when project inherent uncertainty is higher;project planning and control has the same strong effect on process performance, independent of the project inherent uncertainty;the impact of user participation on process performance is found to be significant when project inherent uncertainty is low, but not at high project inherent uncertainty. On the other hand, user participation is the most important factor for product performance. The effects of user participation and intel integration on product performance are stronger when project inherent uncertainty is higher.This study compares the relative importance of ISD successful criterion from the point of view of software developers and users. The findings show that both software developers and users agree that the most important criterion is "meet the requirements of users", followed with "systems with high quality". However, there is a significant difference on the criterion "the systems delivered help to long-term development of user organization". This criterion is ranked second on the users' list, but almost at the low end of the software developers' list.The contribution or innovation in the dissertation may be:1. According to the level of control, this research classifies the risk factors into two categories: the objective uncertainty factors and the subjective behavior factors. Relative to the past categorization scheme based on the risk sources, thiscategorization scheme provides some insights about how to manage a project in the presence of given project uncertainty.In the past, most research focused on in-house development projects or customized projects from the users' perspective. This research studies customized ISD projects from another perspective—system provider. Striving to grasp the nature of multiple sources for uncertainties and dynamic natures for risk of ISD projects, this research analyzes the relationship between project inherent uncertainty, management factors (project planning and control, internal integration, user participation, top management support) and the different dimensions of project performance, developes a two-phase risk based project performance model, and carries out an empirical test. This research provides a new theoretical perspective and systematic framework for the study on ISD project performance.2. Performance risk is defined as the difficulty of estimating project outcomes . Researchers abroad have made several empirical tests on the mediating role of performance risk, but they did not consider the the difference of performance risk in different phases and influential factors on them. Based on the concept of performance risk, this research further defines two variables: performance risk in development phase and performance risk in implementation phase. Supported by the empirical data, it is found that performance risk in development phase and performance risk in implementation phase are two important intervening variables which can explain the effect of project inherent uncertainty and four management factors on the different dimensions of project performance. Moreover, the management factors to influence performance risk in development phase are different from those in implementation phase. Through introducing two-phase risk, the variation of risk and influential factors in the information systems development and implementation process can be dynamically analyzed,which helps to gain inside looks for the risk management strategies in different phases for software development organization.3. A number of IS researchers adopted a contingency perspective of organization theory to study software project risk management. However,research has largely been conceptual and prescriptive in nature, and they did not consider the multi-dimensions of project performance. The results of the empirical research show that when process performance is a key performance criterion, high inherent uncertainty projects call for higher levels of intel integration, and the role of user participation becomes not significant which is different from the low inherent uncertainty group;when product performance is a key performance criterion, high inherent uncertainty projects call forhigher levels of use participation and intel integration. Traditional organization theory often cited a hypothesis that higher levels of uncertainty call for lower levels of formal planning,whereas IS researchers found that high-risk projects call for high levels of formal planning for process performance.However this research finds that project planning and control has the same importance on process performance whether the inherent uncertainty is high or low for temptory organizations such as ISD project.This research extends the contingency theory of IS project management, and have important implications for give some advice for software development otganization to apply proper management strategies according to project type and key performance criterion.4.This research has found some different conclusions in the specific context through the empirical tests on the research model. The finding of this empirical research is that top management support of user organization has not been a direct positive impact on project performance, which has an indirect positive impact on project performance through user participation and performance risk in implementation phase. In addition, the research does not find a significant effect of process performance on product performance, which is different from the conclusion of past empirical research. This result supports the view of some researches that there is a potential conflict between the efficiency of the process and its quality. This result indicates that software project managers should deal with a trade-off relationship between improving process performance and systems quality in the practice of project development.