MSOE Library Home
Login
My List - 0
Help
Home
Search
My Account
Basic
Advanced
Power
History
Search:
Author Keywords
Barcode
Bib No.
Call Number Keyword Search (Enter part of a call number -- use wildcards)
General Keyword Search
Keyword Search of Contents Notes
Keyword Search of Credits Notes (Enter word or words)
Keyword Search of Format
Keyword Search of Item Description
Keyword Search of Publisher's Name
Name Keyword Search
Publication Date Keyword Search
Search Part of an ISBN Number
Search Part of an ISSN Number
Series Title Keywords
Subject Keywords
Title Keywords
Refine Search
> You're searching:
Walter Schroeder Library, Milwaukee School of Engineering
Item Information
Holdings
More by this author
Poquette, Steven E.
Subjects
Technology -- Risk assessment
Computer software -- Development -- Management
Project development
MSEM Thesis.
Browse Catalog
by author:
Poquette, Steven E.
by title:
Applying the paramet...
MARC Display
Applying the parameter method for risk analysis to software project estimates / by Steven E. Poquette.
by
Poquette, Steven E.
Subjects
Technology -- Risk assessment
Computer software -- Development -- Management
Project development
MSEM Thesis.
Description:
167, 11, 4, 3 leaves : ill. ; 29 cm.
Contents:
Advisor: Cecil Head.
Committee members: Dr. Bruce Thompson, Gene Wright.
Introduction -- The case for risk analysis -- Software project risk factors -- Risk analysis methods -- Risk analysis approach -- Risk analysis case study -- Conclusions and recommendations -- Appendix A: Warehouse project risk analysis detail - B: Cost and schedule estimates.
Many times software project managers are required to manage a budget and schedule that was developed very early in the project’s life cycle without much consideration given to potential risk factors. Most software project managers are aware of the risks associated with software development projects, but do not quantify these risks as part of the initial estimate. As a result, senior management is surprised when a software project exceeds the budget and/or the delivery date because of the risk factors associated with the development and implementation. The software project manager needs a tool to identify the probability of delivering the project on, or under budget, and on, or ahead of schedule. The Parameter Method for Risk Analysis, traditionally used in risk analysis for capital projects, is a statistical method that was developed to determine the probability that a capital project will deliver the expected Return On Investment. The proposition of this paper is that the Parameter Method for Risk Analysis can be applied to software project estimates to determine the probability that the software project will exceed the planned project budget and/or schedule. The result of the risk analysis is used by the project manager to develop contingency plans that will reduce the risk, or allow for “what-if” testing. The “what-if” testing and contingency planning identify resource or capital adjustments necessary to achieve the budget and schedule targets. The Parameter Method requires the development of best-case, most likely, and worst-cast estimates for each phase of a project.
From the three estimates a total mean and a total standard deviation are calculated. Both the total mean and the total standard deviation are used to calculate a best-case project estimate, which has a 10% probability of occurring. The range between the best-case and worst-case estimates represents range that covers 80% of the possible estimates. In order to execute the Parameter Method calculations, detailed estimates for each phase of the software project are created, first the initial or most likely estimate, then a best-case (all goes better than planned) estimate, and finally a worst-case estimate (all potential problems occur). In order for the Parameter Method calculations to be useful, a detailed risk analysis must be performed to develop realistic and plausible best-case and worst-case estimates. The Parameter Method is similar to the Program Evaluation Review Technique (PERT) method in that both require the development of three estimates; best-case, most likely and worst-case. The mean calculated from the PERT method is then used as the duration for the activity being estimated. The Parameter Method generates a range of estimates and the estimate that fits the desired confidence level is used as the duration for the activity being estimate. The Parameter Method is used for both cost and schedule estimates, where the PERT method is used just for schedule estimates.
This paper illustrates in detail, the risk factors that are “cost drivers” for software projects. A risk analysis methodology is proposed that identifies and quantifies the risk factors used to develop the best-case and worst-case estimates. The risk factors that the methodology concentrates on are those which have been identify as the most critical for a project; staff availability, staff productivity, staff expertise, completed requirements definition, introduction of new technology and user acceptance of the final system. Cost estimates and the risks associated with computer hardware cost estimates are not addressed in this paper. Software project budget and schedule over-runs are due primarily to unplanned development costs and project delays not accounted for in the original estimate. This paper will focus on the risks associated with software development and installation. Several risk analysis methods are discussed and compared in Chapter 4. The application of each risk analysis method is demonstrated and evaluated. The Parameter Method was selected because it allows a estimate range to be developed where all risk factors can be considered. The risk analysis methodology that provides the inputs to the Parameter Method calculations is designed to identify and quantify the risks associated with a software project. The risk analysis results in developing a best-case, most likely and worst-case person-hour estimate for each phase of the software project. These estimates are then used to develop the project best-case (10% probability), mean (50% probability) and worst-case (90% probability) estimates. The estimates are plotted on a cumulative frequency diagram which can be used to determine the probability of any estimate that lies between the best and worst-case estimates.
The project team and senior management can then determine the risk associated with the original estimate by comparing it to the best, mean, and worst-case estimates. By plotting the original estimate on the cumulative frequency diagram, the project team can determine the probability that the project will be delivered on time and within budget. The risk analysis provides the project team and senior management with a clear picture of the risk associated with the project. Together they can develop alternatives to reduce the risk, or a least there is “buy-in” by senior management as to the risk involved in the project. Senior management can assess the project from a business risk perspective and therefore understand how the project may put the business at risk. The result of performing a risk analysis for a software project, is that the project risks have been identified and their impact estimated. The project manager can develop contingency plans and alternatives in order to reduce the risk. Risk adjusted decisions about staffing, training, resources and development strategies can then be made, before cost and schedule overruns occur.
Copy/Holding information
Location
Collection
Call No.
Status
Walter Schroeder Library
Master's Theses
AC805 .P67 1996
Available
Add Copy to MyList
Format:
HTML
Plain text
Delimited
Subject:
Email to:
Horizon Information Portal 3.25_9382
© 2001-2013
SirsiDynix
All rights reserved.