Software Process and Project Metrics

Study Notes

Software metrics encompasses a variety of measurements aimed at improving software processes and projects.
Measures offer quantitative assessments of attributes related to software products and processes.
Metrics facilitate tasks like estimation, quality control, productivity assessment, and project management.
Indicators, formed from metrics, help provide insights for adjustments in processes and products.

Process Indicators: Offer insights into how well an existing software process is performing.
Project Indicators: Assist project managers by:
- Assessing ongoing project statuses.
- Identifying potential risks.
- Highlighting problem areas before escalation.
- Adjusting workflows and tasks accordingly.
- Evaluating quality control capabilities within the project team.

Metrics are influenced by three factors within a process triangle:
- Skill and motivation of team members.
- Product complexity.
- Software engineering technologies used.
Metrics are also affected by environmental conditions, such as development tools, business constraints, and customer characteristics.

Used to monitor project progress, leveraging historical data for estimates on time and cost.
Helps in minimizing development costs, time, and risks while improving project quality.
Metrics guide project managers in adjusting workflows and assessing product quality regularly.
Examples of project metrics:
- Inputs: Resources required to complete tasks.
- Outputs: Deliverables produced during the process.
- Results: Effectiveness of the deliverables.
Direct measures include cost and effort; indirect measures assess aspects like functionality and maintainability.

Normalized metrics considering the size of the software produced, although not universally accepted as the best evaluation method.
Historical data examples include LOC, defects found, and personnel effort.

Encompasses measures based on the functionality delivered, originally proposed by Albrecht using function points.
Function points are computed based on direct measures of the information domain, including:
- User inputs and outputs.
- User inquiries.
- Logical master files.
- External interfaces.

Correctness: Measured by defects per KLOC regarding requirement conformity.
Maintainability: Assessed through MTTC, evaluating how easily software can be corrected or adapted.
Integrity: Ability to withstand security threats, quantifying security measures.
Usability: Evaluated through user-friendliness based on learning time, efficiency, productivity gains, and subjective user attitudes.