Metrics

Chapter 9
Project Management
Introduction
 Effective project management requires a
well-structured project and diligent
oversight
 A well-structured project consists of a
series of finite, effective, well-defined
tasks
 The phases of a software development
methodology define the tasks to be
managed to some extent
Project Management
Responsibilities
 Establish project schedule
 Establish project budget
 Structure the project into units of work
 Assemble the project team
 Assign units of work to individuals
 Determine necessary resources
 Carry out risk assessment
 Monitor progress of project
 Ensure resulting system meets requirements
Software Metrics
 Reasons to measure software:
– To facilitate estimation of development time
– To assess the productivity of developers
– To assess the quality of the project
 Current schools of thought:
– Size-oriented
– Function-oriented
– Object-oriented
Size-oriented Metrics
 Attempt to quantify software projects by
using the size of the project to normalize
other quality measures
 Possible data to collect:
– number of lines of code
– number of person-months to complete
– cost of the project
– number of pages of documentation
– number of errors corrected before release
– number of bugs found post release
Problems with using Lines
of Code (LOC) as Metric
 Lines of source code comprising a project
are not always good gages to the size and
complexity of a project:
– LOC to complete a task is language dependent
– Code reuse reduces LOC but requires more effort,
thus well-design system are penalized
– Using a LOC based metric encourages
programmers to create more LOC, which is
ultimately less efficient to maintain
Function-Oriented Metrics
 Attempt to measure the functionality of a
software system
 Use a unit of measure called function
point
 Some possible function points:
– Internal data structures
– External data structures
– User inputs
– User outputs
– Transformations
– Transitions
Issues with Using Function-
Oriented Metrics
 Requires that analysis and design of a
project are completed before workload
estimation can occur
 Validity of the workload estimation is
limited to the accuracy of the analysis and
design
 Complexity determination of function
points is subjective
Object-Oriented Metrics
 Suggested measurements for object-
oriented systems:
– Number of scenario scripts
– Number of key classes
– Number of subsystems
 Disadvantages:
– Excludes a history-base of non-object-
oriented projects
Quality Control Metrics
 Correctness
– Defects per thousand LOC
 Maintainability
– Mean time to change
 Integrity
– Likelihood of thwarting an attack
 Usability
– Skill required of users
– Time required to become proficient
– Net increase in productivity
– Users’ attitude toward system
Other Project Management
Concepts
 Mythical staff-month
 Configuration management
 Change control
 Configuration Audit
 Configuration status reporting
Project Planning
 Project planning requires:
– Defining the iterations of the project
– Specifying subtasks
– Determining the project schedule and allocating
time for each subtask
– Associating deliverables with each subtask to
verify progress
– Dividing the subtasks among the developers
– Scheduling any interdependent tasks to minimize
delays - use task network or PERT chart
Monitoring Project Progress
 Develop project milestones with
associated deliverables to gage progress
 Milestones should be created so that the
project manager receives sufficient
feedback at regular intervals
 The feedback should take the form of a
natural artifact of the development
process
 See figure 9.9 for a list of deliverables
Four Stages of Team
Development
 Forming
 Storming
 Norming
 Performing
Conflict Resolution
Strategies
 Arbitration
 Rules and regulations
 Confrontation
 Negotiation
 Separation
 Neglect
 Coordination device
Risk Management
 Risk management provides a structured
evaluation of a development project to draw
attention to sources of risk
 The need for risk management is demonstrated
by the high failure rate for large-scale software
development initiatives
 Successful project management relies on the
additional time that is built into the development
schedule to accommodate some level of delay
due to risk factors
What is Risk?
 A risk is any unanticipated condition or
event that causes one or more tasks to be
delayed, lengthened, or fail
 Risks can delay or prevent the completion
of a task or project as a whole
 Two very general categories of risk will be
identified here, technical and human risk
Sources of Technical Risk
 Project complexity
 Project size
 Use of state-of-the-art technology
 Network vulnerability
 Disgruntled employees
 Potential for white-collar crime
 Data attainability
 Accuracy of data source
 Need for high-quality graphics
Sources of Human Risk
 Development team  Administration
– Productivity – Budgetary
– Experience constraints
– Knowledge – Project priority
– Dedication – Realistic
 End users expectations
– Technical knowledge
– Support for project
– Agreement on system
Consequences of Risk
 Delay project
 Compromise the quality of the
project
 Cause the project to fail
 Cause the project to be too
expensive to implement or run
Reducing Risk
 Early project evaluation
 Early implementation of risky system
aspects
 Early use of new technology
 Early resolution of class interaction
problems

Metrics

Uploaded by

Copyright:

Available Formats

Metrics

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Metrics

Uploaded by

Copyright:

Available Formats

Chapter 9

You might also like