PadhaikoAll subjects

Project Quality Management

ICT Project Management (ENCT 412)

Open in the Padhaiko app

9.1 Quality theories

In ICT project management, delivering software on time and within budget fails if the system crashes in production, violates security policies, or does not satisfy user needs. Project Quality Management ensures that the project satisfies the needs for which it was undertaken. According to ISO 9000, quality is "the degree to which a set of inherent characteristics fulfills requirements." Crucially, quality (meeting stated/implied requirements) differs from grade (a category assigned to deliverables having the same functional use but different technical characteristics—low grade is acceptable if appropriate, whereas low quality is never acceptable).

Definition

Foundational Quality Theorists and Principles

  1. W. Edwards Deming: Advocated continuous improvement via the Plan–Do–Check–Act (PDCA) cycle, Statistical Process Control (SPC), and his 14 Points for Management. Argued that roughly of quality problems stem from management-controlled systems rather than individual workers.
  2. Joseph M. Juran: Defined quality as "fitness for use" (conformance to user expectations, not merely specifications). Proposed the Juran Quality Trilogy: Quality Planning, Quality Control, and Quality Improvement, and popularized the Pareto Principle (80/20 rule) in quality.
  3. Philip B. Crosby: Coined "Quality is Free"—preventing defects costs less than detecting and fixing them later. Defined quality as conformance to requirements and promoted Zero Defects ("do it right the first time").
  4. Kaoru Ishikawa: Pioneered Quality Circles and the Cause-and-Effect (Fishbone / Ishikawa) Diagram to systematically trace root causes of defects.
  5. Genichi Taguchi: Introduced the Taguchi Loss Function (any deviation from the target value causes progressive loss to society, even inside tolerance limits) and robust design methods.

Modern ICT quality management builds on these theories through:


9.2 Quality planning and requirements, cost of quality, quality management plan

Plan Quality Management is the process of identifying quality requirements and standards for the project and its deliverables, and documenting how the project will demonstrate compliance.

9.2.1 Quality Planning and Requirements

In ICT projects, quality requirements must be converted into measurable, testable acceptance criteria before design and coding begin. Common ICT quality attributes include:

9.2.2 Cost of Quality (CoQ)

Cost of Quality (CoQ) is the total cost incurred over the product life cycle to prevent nonconformance, appraise deliverables, and repair failures when requirements are not met:

Categories of Cost of Quality (CoQ) in ICT Projects
Major GroupSub-categoryPurposeICT Project Examples
Cost of ConformancePrevention CostsAvoid defects before they occurDeveloper training, coding standards, architecture reviews, automated CI/CD setup
Cost of ConformanceAppraisal CostsAssess deliverables to detect defectsUnit/integration testing, static code analysis, peer code reviews, quality audits
Cost of NonconformanceInternal Failure CostsFix defects found before customer releaseDebugging, rework, regression re-testing, wasted sprint capacity
Cost of NonconformanceExternal Failure CostsFix defects found after customer releaseProduction hotfixes, SLA penalty payouts, help-desk surge, lost customer trust

9.2.3 Quality Management Plan

The Quality Management Plan (QMP) is a subsidiary component of the Project Management Plan that defines how quality policies will be implemented:

  1. Quality Objectives & Standards: Applicable IEEE/ISO standards, regulatory rules, and service-level targets.
  2. Quality Roles and Responsibilities: Clear ownership across the Project Manager, QA Lead, Developers, and Business Analysts.
  3. Quality Metrics & Tolerance Thresholds: Quantifiable metrics (e.g., defect density , test coverage ).
  4. QA and QC Activities: Schedules for peer reviews, audits, test levels, and release gates.
  5. Defect Management & Escalation: Severity/priority classification, bug triage workflow, and Corrective and Preventive Action (CAPA) procedures.

9.3 Quality assurance, quality audit, approach to a quality audit

Manage Quality (Quality Assurance, QA) is the process of translating the Quality Management Plan into executable quality activities that incorporate the organization's quality policies into the project.

AspectQuality Assurance (QA)Quality Control (QC)
Primary FocusProcess-oriented: builds quality into workflowsProduct-oriented: inspects finished deliverables
NatureProactive (defect prevention)Reactive / detective (defect detection)
TimingContinuous throughout project executionAt specific checkpoints and deliverable completions
Typical TechniquesQuality audits, process checklists, design/code standards, trainingTesting, inspection, control charts, Pareto charts, walkthroughs
OutputProcess improvements, audit reports, updated standardsVerified deliverables, defect logs, change requests

9.3.1 Quality Audit

Definition

Quality Audit

A quality audit is a structured, independent process to determine whether project activities comply with organizational and project policies, processes, and procedures. Audits may be scheduled or random, and are conducted by internal audit teams or external third-party auditors.

Objectives of a quality audit include:

9.3.2 Approach to a Quality Audit

An effective ICT quality audit follows five structured phases:

  1. Audit Planning and Scoping: Define audit objectives, scope (e.g., release pipeline, SDLC documentation, test traceability), criteria (ISO 9001, QMP), audit team, and schedule.
  2. Preparation and Document Review: Distribute an audit checklist, review the QMP, SOPs, configuration logs, and previous audit findings before meeting the project team.
  3. Audit Execution (Fieldwork): Conduct opening meeting, interview developers/testers, sample work artifacts (e.g., pull request approvals, test execution logs, requirement traceability matrices), and verify actual practice against documented procedures.
  4. Reporting and CAPA: Categorize findings into Major Nonconformities, Minor Nonconformities, and Observations; present the audit report; and require the project team to submit Corrective and Preventive Actions (CAPA) with root-cause analysis and due dates.
  5. Follow-Up and Closure: Re-verify implemented corrective actions on the target dates and formally close the audit items.

9.4 Quality control process, control chart, Pareto charts

Control Quality (QC) is the process of monitoring and recording results of executing quality management activities to assess performance and ensure project outputs are complete, correct, and meet customer expectations.

9.4.1 Quality Control Process

The QC process takes deliverables, work performance data, and quality metrics as inputs; applies inspection and statistical tools (such as Ishikawa's Seven Basic Quality Tools: cause-and-effect diagram, flowchart, checksheet, Pareto chart, histogram, control chart, and scatter diagram); and produces verified deliverables, quality control measurements, and change requests for defect repair or rework.

9.4.2 Control Charts

A control chart is a time-ordered graphic display of process performance used to determine whether a process is statistically stable or has predictable performance.

Definition

Control Chart Limits and Out-of-Control Rules

For a process with mean (estimated by sample average ) and process standard deviation :

A process is out of control (exhibits special-cause variation) if:

  1. Any single data point falls outside the Upper Control Limit () or Lower Control Limit ().
  2. Rule of Seven: Seven (or more) consecutive data points lie entirely on one side of the center line, or show a continuous upward or downward trend—even if all points remain inside the control limits.

9.4.3 Pareto Charts

A Pareto chart is a specialized vertical bar chart combined with a cumulative percentage line, named after economist Vilfredo Pareto and popularized in quality by Joseph Juran (80/20 rule).

Example 1

1

During a 6-month web-application project, the team recorded the following quality-related expenses:

Classify each item into the four CoQ categories, and calculate: (a) Cost of Conformance, (b) Cost of Nonconformance, (c) Total Cost of Quality, and (d) the percentage share of conformance vs. nonconformance costs.

Classifying Each Cost Item:

  • Prevention Cost: Secure-coding training and CI setup
  • Appraisal Cost: Integration testing and code reviews
  • Internal Failure Cost: Pre-release bug fixing and re-testing
  • External Failure Cost: Post-release hotfixes and SLA penalties

Computing Cost of Conformance, Nonconformance, and Total CoQ:

Percentage Breakdown:

Over of total quality expenditure went to fixing failures—indicating under-investment in upfront prevention and appraisal.

Answer:
  • (a) Cost of Conformance: ().
  • (b) Cost of Nonconformance: ().
  • (c) Total Cost of Quality: .
  • (d) Share: Conformance accounts for and Nonconformance for .
Example 2

2

  1. An automated nightly build pipeline has a historical mean build time and standard deviation . Calculate the control limits ( and ), and state whether a sequence of consecutive builds taking indicates the process is in control.
  2. In the same release, production defects were categorized as: UI Alignment (), API Timeout (), Validation Error (), and Export Formatting (). Construct a Pareto ranking and determine which categories constitute the vital .

Part 1 — Control Limits and Stability Check:

All observations lie inside ; however, all consecutive points exceed the mean . By the Rule of Seven, seven or more consecutive points on one side of the center line signal a special-cause shift—the build process is out of control and requires investigation.

Part 2 — Pareto Ranking and Cumulative Percentages: Sorting defect categories in descending order of frequency ( total):

Pareto Analysis of Production Defects
RankDefect CategoryCountShare ()Cumulative CountCumulative ()
API Timeout
Validation Error
Export Formatting
UI Alignment

The top two categories (API Timeout and Validation Error) account for out of defects, or exactly of total defects.

Answer:
  1. , . Despite lying within limits, the process is out of control because consecutive points fall above (violating the Rule of Seven).
  2. API Timeout () and Validation Error () together form the vital of defects in the Pareto ranking.

9.5 Testing of IT system, the test life cycle

Testing is the primary Quality Control mechanism in ICT projects. Rather than proving the absence of bugs, systematic testing verifies conformance to functional and non-functional requirements and measures release readiness.

9.5.1 Levels and Types of IT System Testing

Levels of IT System Testing
Testing LevelScope & PurposeTypical Owner
Unit TestingVerifies individual functions, classes, or modules in isolation (often automated with mocks/stubs)Developers
Integration TestingVerifies interfaces, APIs, and data flow between combined modules or external servicesDevelopers / QA
System TestingEvaluates the complete, integrated system end-to-end against functional and non-functional SRS requirementsIndependent QA Team
User Acceptance Testing (UAT)Validates business workflows and usability in a realistic environment before final sign-offClient / End Users

Beyond functional levels, ICT systems undergo Regression Testing (re-running tests after code changes to verify nothing broke), Performance/Load Testing (measuring response time and concurrency limits), Security Testing (penetration and vulnerability scanning), and Usability Testing.

9.5.2 The Software Testing Life Cycle (STLC)

The Software Testing Life Cycle (STLC) defines a systematic sequence of phases executed by the testing team alongside the SDLC:

Definition

Phases of the Software Testing Life Cycle (STLC)

  1. Requirement Analysis: QA reviews SRS and user stories to identify testable requirements, clarify ambiguities, and produce a Requirement Traceability Matrix (RTM).
  2. Test Planning: Test Lead prepares the Test Plan, defining test scope, strategy, schedule, resource needs, tools, risks, and entry/exit criteria.
  3. Test Case Development: Testers author detailed test cases, test scenarios, and test data (and automation scripts) mapped to requirements.
  4. Test Environment Setup: Provisioning hardware, OS, database seeds, network/sandbox endpoints, and executing smoke tests to confirm environment readiness.
  5. Test Execution: Executing test cases, logging actual vs. expected results, filing defects in a bug tracker, and performing re-testing and regression testing as fixes arrive.
  6. Test Cycle Closure: Verifying exit criteria (e.g., zero Severity-1 open defects, pass rate ), archiving test artifacts, computing Defect Removal Efficiency (), and delivering the Test Summary Report.

A standard quantitative metric computed at test closure is Defect Removal Efficiency ():

where is the number of defects detected during internal testing before release and is the number of defects found by users after release.

Example 3

3

During Release 1.0 of a banking portal, the QA team detected defects prior to launch, while customers reported defects in the first month of production. After introducing a formal Quality Audit and automated STLC integration gates for Release 2.0, internal QA caught defects and only defects escaped to production.

  1. Compute the Defect Removal Efficiency () for Release 1.0 and Release 2.0.
  2. Outline the five-phase Quality Audit approach the QA Lead used between releases to improve STLC effectiveness.

Computing Defect Removal Efficiency ():

  • Release 1.0:
  • Release 2.0:

The process improvements increased by percentage points and reduced production escapes by .

Five-Phase Quality Audit Approach Applied Between Releases:

  1. Audit Planning: Scoped the audit to Requirement Traceability Matrix coverage and UAT sign-off rules.
  2. Preparation: Reviewed the Quality Management Plan, test plan, and Release 1.0 defect logs using an audit checklist.
  3. Fieldwork: Interviewed QA engineers and inspected sample test runs, discovering unverified API edge cases.
  4. Reporting & CAPA: Logged a Major Nonconformity for missing integration checks and issued CAPA requiring automated regression gates before UAT.
  5. Follow-Up & Closure: Verified CI/CD test gates were active for Release 2.0 and closed the audit findings.
Answer:
  1. and .
  2. The five-phase quality audit (Planning Preparation Fieldwork Reporting & CAPA Follow-Up & Closure) identified missing integration test gates and enforced automated regression checks before UAT.

Exercises

Practice Exercises — Project Quality Management

Question 1

Summarize the key quality contributions of W. Edwards Deming, Joseph Juran, and Philip Crosby. Why does ISO 9000 distinguish between quality and grade?

Contributions of Quality Theorists:

  • W. Edwards Deming: Promoted the PDCA (Plan–Do–Check–Act) cycle, Statistical Process Control, and his 14 Points, emphasizing that management owns roughly of quality issues.
  • Joseph Juran: Defined quality as "fitness for use," created the Quality Trilogy (Planning, Control, Improvement), and applied the Pareto (80/20) principle.
  • Philip Crosby: Defined quality as conformance to requirements, championed Zero Defects, and argued "quality is free" because prevention costs less than rework.

Quality vs. Grade: Quality measures how well inherent characteristics fulfill requirements, whereas grade categorizes functional design features (e.g., basic vs. enterprise software tier). A low-grade product with zero bugs has high quality, whereas a high-grade product full of crashes has low quality.

Answer:

Deming (PDCA, SPC, management ownership), Juran (fitness for use, Quality Trilogy), and Crosby (conformance to requirements, Zero Defects, "quality is free") laid the foundation of modern quality management. Quality is fulfillment of requirements (never compromised), while grade is feature richness (acceptable at any agreed level).

Question 2

Differentiate between Quality Assurance (QA) and Quality Control (QC), and explain the step-by-step approach to conducting a quality audit in an ICT project.

QA vs. QC:

  • QA is process-oriented and proactive, verifying that teams follow defined standards and procedures (e.g., audits, peer-review policies) to prevent defects.
  • QC is product-oriented and detective, inspecting specific deliverables (e.g., system testing, control charts, Pareto charts) to detect defects and verify correctness.

Five-Step Approach to a Quality Audit:

  1. Plan & Scope: Establish audit criteria, scope, team, and schedule.
  2. Prepare: Review the QMP, SOPs, and prior records; finalize an audit checklist.
  3. Execute Fieldwork: Conduct interviews and sample project artifacts against procedures.
  4. Report & CAPA: Document nonconformities/observations and assign Corrective and Preventive Actions (CAPA).
  5. Follow-Up & Close: Verify CAPA completion and formally close findings.
Answer:

QA proactively audits and improves processes, whereas QC inspects specific deliverables for conformance. A quality audit proceeds through Planning, Preparation, Fieldwork, Reporting/CAPA, and Follow-Up/Closure.

Question 3

An IT project spends on developer training and architecture reviews, on automated test execution, fixing bugs caught during QA, and handling post-launch customer patches and SLA penalties. Calculate the Cost of Conformance, Cost of Nonconformance, Total Cost of Quality, and the percentage of Total CoQ caused by nonconformance.

Cost of Conformance (Prevention + Appraisal):

Cost of Nonconformance (Internal Failure + External Failure):

Total CoQ and Nonconformance Percentage:

Answer:

Cost of Conformance , Cost of Nonconformance , Total CoQ , and Nonconformance represents of Total CoQ.

Question 4
  1. Response latency of a microservice has a historical mean and standard deviation . Compute the Upper Control Limit (UCL) and Lower Control Limit (LCL). Explain what happens if a new sample records while the customer SLA ceiling is .
  2. During testing of the microservice, defects were detected before deployment and defects were reported in production. Compute the Defect Removal Efficiency ().

Part 1 — Control Limits and Control vs. Specification Analysis:

Even though satisfies the customer specification limit (), it exceeds , meaning the process is statistically out of control (special-cause variation is present and must be investigated before the process degrades further).

Part 2 — Defect Removal Efficiency ():

Answer:
  1. , . A reading of is out of statistical control () even though it meets the SLA specification.
  2. .
Question 5

Describe the six phases of the Software Testing Life Cycle (STLC) and explain how Unit, Integration, System, and User Acceptance Testing (UAT) fit into an ICT project's release workflow.

Four Testing Levels in Sequence:

  • Unit Testing: Developers verify individual modules/functions in isolation.
  • Integration Testing: Verifies data exchange and API contracts between combined modules.
  • System Testing: QA validates the complete integrated application against all functional and non-functional SRS requirements.
  • User Acceptance Testing (UAT): End users/clients validate real business workflows to grant release sign-off.

Six Phases of the STLC:

  1. Requirement Analysis: Identify testable requirements and build the Requirement Traceability Matrix (RTM).
  2. Test Planning: Define scope, tools, schedule, and entry/exit criteria in the Test Plan.
  3. Test Case Development: Design test cases, scripts, and test data.
  4. Test Environment Setup: Configure servers, databases, and run smoke tests.
  5. Test Execution: Run tests, log defects, re-test fixes, and run regression suites.
  6. Test Cycle Closure: Evaluate exit criteria, compute metrics (, pass rate), and issue the Test Summary Report.
Answer:

Testing progresses from Unit Integration System UAT across six STLC phases: Requirement Analysis, Test Planning, Test Case Development, Test Environment Setup, Test Execution, and Test Cycle Closure.

← Project Cost ManagementProject Communication Management →