Unit 2: Software metrics and quality assurance

• Measurement: assigning numbers or symbols to attributes of entities according to defined rules.

• Good metrics are simple, computable, consistent, objective, language-independent and give useful feedback.

• Customer satisfaction metrics: surveys on a five-point scale (very satisfied … very dissatisfied); measures such as percentage satisfied and net satisfaction index; CUPRIMDSO (capability, usability, performance, reliability, installability, maintainability, documentation, service, overall).
• Function points: measure functionality from the user's view — inputs, outputs, inquiries, internal files and external interfaces weighted by complexity, adjusted by 14 factors; useful because they are independent of language. FP = count total × (0.65 + 0.01 × ΣFi).

Software quality is conformance to explicit requirements, standards and implicit expectations. It has two views: quality of design and quality of conformance. Cost of quality: prevention costs (training, planning), appraisal costs (reviews, testing) and failure costs (internal — before release; external — after release, the most expensive).

SQA is a planned set of activities to ensure software meets quality standards: setting standards, reviews and audits, testing, error collection and analysis, change management and education. A formal technical review (FTR) is a meeting of 3–5 people who examine a work product (design or code) to find errors early. Guidelines: review the product, not the producer; set an agenda; limit debate; keep it under two hours; record issues. Statistical SQA collects defect data, traces each defect to its cause and uses the Pareto principle (80% of defects come from 20% of causes) to fix the vital few causes first. Six Sigma aims for 3.4 defects per million opportunities (DMAIC: Define, Measure, Analyse, Improve, Control).

• McCall's model groups factors into product operation (correctness, reliability, efficiency, integrity, usability), product revision (maintainability, flexibility, testability) and product transition (portability, reusability, interoperability).

• Advantages of SQC: detects problems early, reduces inspection cost, improves quality and productivity, provides objective data, builds quality consciousness.
• Limitations: needs trained staff and accurate data; detects but does not itself remove causes; assumes stable measurement; may be costly for small runs.