This preview shows page 1. Sign up to view the full content.
Unformatted text preview: like minutes to do a task)
• the ± uncertainty borders of the estimate.
• the % level of impacts with respect to primary targets and deadlines (100% means
meets target on time)
• the credibility level of the estimates and uncertainty (scale 0.0 to 1.0)
which is derived from the parameters
• experience data for making the estimates (facts, case studies, other projects)
• the sources of the experience data (people, groups, papers, studies) 5. Use metrics to describe solutions, designs, and architecture.
All ‘designs’ have multiple performance/quality/cost attributes, that define ‘how well’ the
designs satisfy our requirements.
In my view it should be normal practice to evaluate all major impact designs, strategies and
architectures, against all numeric critical performance and cost targets.
Tools like QFD (Quality Function Deployment) are trying ‘structurally’, but in my view are not
good enough at defining objective real world metrics [Gilb QFD]. Giving a subjective impact
evaluation a number like 5, 3 or 7 is not real world metrics – and that is but one small example of
my complaint against QFD and similar weak practices. As a consequence, I personally have no
View Full Document
This document was uploaded on 02/26/2014 for the course E 515 at University of Louisiana at Lafayette.
- Spring '13
- Systems Engineering