Quality Management and Statistical Process Control

Introduction

Quality management is the systematic direction and control of activities that determine whether products, services and processes meet specified requirements and satisfy customers. In contemporary operations, quality is not restricted to inspection at the end of production. It is designed into processes, measured with evidence, improved through structured methods and sustained by organisational culture.

Quality has several dimensions. In manufacturing these may include performance, conformance, reliability, durability, serviceability, aesthetics and perceived quality. In services, relevant dimensions include reliability, responsiveness, assurance, empathy and tangibles. A product can conform to its technical specification yet fail because it is unavailable, difficult to use or supported poorly. Operations managers therefore need both technical and customer-oriented approaches.

This chapter examines Total Quality Management (TQM), the cost of quality, ISO 9001, the seven basic quality tools, Statistical Process Control (SPC), process capability, acceptance sampling, Six Sigma DMAIC and customer-focused service quality.

Learning Objectives

By the end of this chapter, the learner should be able to:

1. Explain quality management, TQM and the prevention-based quality philosophy.
2. Classify and calculate the cost of quality.
3. Apply the principles and process logic of ISO 9001.
4. Select and use the seven basic quality tools.
5. Construct and interpret appropriate control charts.
6. Evaluate process capability under stated assumptions.
7. Distinguish SPC from acceptance sampling and use sampling results responsibly.
8. Apply the Six Sigma DMAIC framework to operational problems.
9. Evaluate service quality from the customer’s perspective.
10. Discuss implementation, ethical, technological and governance issues in quality management.

1. From Inspection to Total Quality Management

Traditional quality control relied heavily on inspection: defective items were identified after production. Inspection remains useful, but it is a detection activity. It does not by itself remove the causes of defects. Modern quality management emphasises prevention, process ownership and continual improvement.

The central principle is that output quality is largely determined by process quality. If a process is unstable, increasing final inspection may increase cost without reliably improving performance. Quality should therefore be built into product design, supplier selection, work methods, equipment maintenance, training and information systems.

Important contributions include:

• Walter Shewhart: developed statistical process control and distinguished common-cause from special-cause variation.
• W. Edwards Deming: emphasised systems thinking, reduction of variation, leadership responsibility and the Plan–Do–Study–Act cycle.
• Joseph Juran: described quality as fitness for use and highlighted quality planning, control and improvement.
• Philip Crosby: promoted conformance to requirements and prevention, associated with the phrase “quality is free,” meaning that prevention can reduce the larger costs of failure.
• Kaoru Ishikawa: developed the cause-and-effect diagram and stressed company-wide quality participation.

Total Quality Management

TQM is an organisation-wide management philosophy in which every function, process and employee contributes to customer satisfaction and continual improvement. Its characteristic principles are:

1. Customer focus: requirements are identified, translated into specifications and monitored through customer feedback.
2. Leadership: managers establish purpose, resources, priorities and an environment in which problems can be reported.
3. Employee involvement: those closest to the work participate in problem solving and improvement.
4. Process approach: work is managed as interconnected processes with inputs, activities, controls and outputs.
5. Systems thinking: quality problems often cross departmental boundaries.
6. Evidence-based decisions: data, not anecdote or blame, guides action.
7. Continual improvement: performance is improved incrementally or through major redesign where necessary.
8. Mutually beneficial supplier relationships: suppliers are treated as partners in controlling upstream variation.

TQM is not simply a collection of tools. It requires alignment between strategy, structure, behaviour, measurement and rewards. A manager who demands high throughput while penalising employees for stopping a defective process creates a conflict that undermines TQM.

2. Cost of Quality

The cost of quality is the total cost associated with preventing poor quality, evaluating conformity and dealing with failures. It is not the cost of producing quality itself; rather, it describes the economic consequences of the quality system.

The basic relationship is:

where COQ is total cost of quality, C_P is prevention cost, C_A is appraisal cost, C_IF is internal failure cost and C_EF is external failure cost. The categories are accounting classifications and must be defined consistently within the organisation.

Worked Example: Cost of Quality

A factory reports annual prevention costs of $40,000, appraisal costs of $60,000, internal failure costs of $90,000 and external failure costs of $110,000. Therefore:

Prevention and appraisal costs equal $100,000, while failure costs equal $200,000. Management should not automatically cut appraisal costs. A sound decision compares the expected reduction in failure costs with the additional prevention or appraisal investment. Excessive inspection can also conceal weak processes.

Quality costs can be expressed as a percentage of sales:

If annual sales are $6,000,000, the percentage is 5%. Comparisons are meaningful only when the accounting boundaries, product mix and sales basis are stable.

3. ISO 9001 and the Quality Management System

ISO 9001 specifies requirements for a quality management system (QMS). It does not certify that every product is perfect, nor does it prescribe one production method. Certification provides evidence that the organisation has a documented and controlled system for meeting requirements and improving its processes.

The ISO 9001 framework is based on seven quality management principles:

1. Customer focus
2. Leadership
3. Engagement of people
4. Process approach
5. Improvement
6. Evidence-based decision making
7. Relationship management

A QMS typically includes organisational context, interested-party requirements, leadership responsibilities, quality objectives, risk-based planning, operational controls, competence, documented information, performance evaluation and improvement.

The process approach can be represented as inputs → activities → outputs, with resources, controls, measures and feedback. A process owner is accountable for performance even when activities cross several departments. The Plan–Do–Check–Act logic is particularly useful:

• Plan: understand requirements, risks, resources and objectives.
• Do: operate the process under controlled conditions.
• Check: monitor, measure, audit and analyse results.
• Act: correct causes and improve the system.

ISO 9001 in Practice

Implementation should begin with the organisation’s actual processes rather than with documentation for its own sake. Useful steps are:

1. Map key processes and identify customers, suppliers, inputs and outputs.
2. Identify legal, regulatory, contractual and customer requirements.
3. Assess risks and opportunities.
4. Establish measurable quality objectives.
5. Define responsibilities, competence requirements and controls.
6. Maintain controlled documented information.
7. Audit processes and review performance.
8. Correct nonconformities and verify corrective-action effectiveness.

Audits should be independent enough to provide credible evidence. A nonconformity should be supported by an objective finding, linked to a requirement and addressed through root-cause analysis. Certification must never be treated as a substitute for operational learning.

4. The Seven Basic Quality Tools

The seven tools are simple, but their value depends on correct problem definition, reliable data and disciplined interpretation.

4.1 Check Sheet

A check sheet is a structured form for collecting data at the location where events occur. Categories, time periods and recording rules must be defined in advance. It is useful for defect counts, equipment stoppages, complaint types and location-based observations.

4.2 Histogram

A histogram groups numerical observations into intervals and displays their frequencies. It helps reveal centre, spread, skewness, gaps and possible multiple populations. The number and width of intervals influence appearance, so interpretation should be supported by raw data and process knowledge.

4.3 Pareto Chart

A Pareto chart ranks categories from largest to smallest, usually with a cumulative-percentage line. It supports prioritisation, but the “vital few” are not guaranteed to represent exactly 20% of causes. Categories must be defined consistently, and high-frequency problems should also be evaluated for severity and risk.

4.4 Cause-and-Effect Diagram

The fishbone diagram organises possible causes under headings such as people, machines, methods, materials, measurement and environment. It is a hypothesis-generation tool, not proof of causation. Suspected causes must be tested with data.

4.5 Scatter Diagram

A scatter diagram displays paired observations to investigate association between two variables, such as temperature and defect rate. Correlation does not establish causation. A third variable, restricted data range or time sequence may explain an apparent relationship.

4.6 Flowchart

A flowchart represents process sequence, decisions, rework loops and handoffs. It often exposes duplicate approvals, unclear ownership and unnecessary movement. A process map should reflect actual practice, not merely the intended procedure.

4.7 Control Chart

A control chart plots observations over time with a central line and statistically derived control limits. It distinguishes routine process variation from unusual signals. It is discussed in detail below.

5. Statistical Process Control

SPC applies statistical methods to monitor process behaviour and support timely action. Two sources of variation are fundamental:

• Common causes: natural, persistent sources within the process system, such as normal machine variation or standard material differences.
• Special causes: identifiable and unusual sources, such as a broken tool, incorrect setting or contaminated batch.

A stable process has only common-cause variation, although it may still be incapable of meeting specifications. Stability and capability are different questions.

Control Limits and Specification Limits

Control limits are calculated from process data and indicate expected statistical behaviour. Specification limits are set by customers, designers, regulators or contracts. Control limits answer “Is the process behaving consistently?” Specification limits answer “Is the output acceptable?” Confusing these concepts can lead to incorrect decisions.

Variables Control Charts

For quantitative measurements, an X̄-R chart is commonly used for rational subgroups. For subgroup size n, subgroup mean is:

where X_ij is observation j in subgroup i. The average of subgroup means is X̄̄, and the average subgroup range is R̄. Control limits are commonly written:

For the range chart:

The constants A₂, D₃ and D₄ depend on subgroup size and assume approximately independent observations from a stable process. For an X̄-R chart, the range estimates short-term within-subgroup variation; constants must be selected for the actual subgroup size.

Attribute Charts

A p chart monitors the fraction nonconforming when sample sizes may vary. If d_i nonconforming units are found in sample n_i, then p_i = d_i/n_i. With a constant or estimated process fraction p̄, limits are:

The lower limit is set to zero if the calculated value is negative. These limits assume independent Bernoulli outcomes and are approximate for sufficiently large samples. For varying sample sizes, use n_i for each point. A c chart counts defects when the opportunity area is constant; a u chart monitors defects per unit when area or sample size varies.

Interpreting Signals

Potential signals include a point beyond a control limit, a sustained run on one side of the centre line, a trend, cyclic behaviour or an unusual pattern. Rules should be defined before analysis to avoid selectively interpreting results. When a signal occurs, preserve evidence, investigate the process, identify the cause and document the action. Do not adjust a stable process merely because an observation is not exactly at the centre line; unnecessary adjustment can increase variation.

6. Process Capability

Capability compares the natural process spread and location with specification requirements. It should be assessed only after the process is demonstrated to be statistically stable and the measurement system is adequate.

For a stable approximately normal process with lower specification limit LSL, upper specification limit USL, mean μ and standard deviation σ, the capability indices are:

C_p measures potential capability if the process is centred; C_pk also accounts for off-centre operation. If specifications are one-sided, an appropriate one-sided index should be used rather than forcing a two-sided formula.

Worked Example: Capability

Suppose a shaft diameter has specifications of 49.5 to 50.5 mm, process mean 50.1 mm and standard deviation 0.10 mm. Then:

The process has greater potential than actual capability because it is closer to the upper specification limit. A manager may first centre the process, but should not assume that capability indices alone prove conformance. Non-normal distributions, autocorrelation, unstable processes and biased measurement systems require additional analysis.

A measurement-system analysis, often including repeatability and reproducibility studies, is important. If measurement error is large relative to process variation or specification width, capability conclusions may be misleading.

7. Acceptance Sampling

Acceptance sampling is a decision procedure in which a sample is taken from a lot and the lot is accepted or rejected according to a rule. It is not the same as SPC. SPC controls an ongoing process; acceptance sampling makes a disposition decision about a lot.

A single-sampling plan is described by sample size n and acceptance number c. For a sample of n units, the lot is accepted if the number of nonconforming units is less than or equal to c, otherwise it is rejected. Sampling reduces inspection effort but does not guarantee that every unit in an accepted lot is conforming.

For a lot with defect probability p, the probability of observing exactly x nonconforming units in a sample of size n, when a binomial model is appropriate, is:

The probability of accepting the lot under a single-sampling plan is:

This assumes independent sampling and a sufficiently large lot relative to the sample, or sampling with replacement as an approximation. For finite lots without replacement, a hypergeometric model may be more appropriate.

An Operating Characteristic (OC) curve shows the probability of acceptance for different quality levels. A good plan protects both producer and consumer interests by specifying risks. Producer’s risk is rejecting a lot of acceptable quality; consumer’s risk is accepting a lot of unacceptable quality. Acceptance sampling should be combined with supplier development and process improvement, not used to justify chronic poor quality.

8. Six Sigma and DMAIC

Six Sigma is a data-driven approach to reducing variation and defects while improving process performance. Its projects commonly use DMAIC:

• Define: state the customer, problem, scope, business case, objectives and critical-to-quality requirements.
• Measure: define operational measures, validate the measurement system and establish baseline performance.
• Analyze: identify and verify root causes using process knowledge and data.
• Improve: develop, test and implement solutions; use designed experiments where appropriate.
• Control: standardise the improved process, monitor performance and transfer ownership.

A project charter prevents scope drift. A SIPOC diagram—suppliers, inputs, process, outputs and customers—provides a high-level boundary. Critical-to-quality characteristics translate the voice of the customer into measurable requirements.

Defects per million opportunities (DPMO) can be calculated as:

For example, 18 defects are found in 1,200 forms, each with 3 possible defect opportunities:

DPMO is meaningful only when a defect opportunity is defined consistently. Converting DPMO to a “sigma level” requires a stated convention; published sigma tables may use different assumptions, including a long-term shift. Managers should therefore report the underlying defect rate and assumptions rather than relying on a label alone.

Lean and Six Sigma are complementary. Lean focuses strongly on flow and waste reduction, while Six Sigma focuses on variation and defect reduction. Both can fail if projects are selected for statistical elegance rather than customer or strategic value.

9. Customer-Focused Service Quality

Service quality is difficult to manage because services are often intangible, produced and consumed simultaneously, variable across employees and locations, and affected by customer participation. The customer evaluates not only the outcome but also the process of delivery.

A practical service-quality framework considers:

• Reliability: performing the promised service accurately.
• Responsiveness: providing prompt help and information.
• Assurance: demonstrating competence, courtesy, credibility and safety.
• Empathy: understanding and giving individual attention.
• Tangibles: facilities, equipment, appearance and communication materials.

The SERVQUAL tradition compares customer expectations with perceptions across these dimensions, but organisations should adapt measures to their context rather than treating a questionnaire as a complete truth. Useful operational measures include first-contact resolution, waiting time, abandonment rate, complaint recurrence, on-time service completion and customer effort.

A service blueprint maps customer actions, front-stage employee actions, back-stage activities and support processes. Failure points become candidates for mistake-proofing, standard work, staff training or technology redesign. Complaint data should be treated as process information, not merely as a public-relations problem.

Quality measures should combine lagging indicators, such as complaints and refunds, with leading indicators, such as training completion, response time and process adherence. Excessive standardisation can damage empathy, while excessive discretion can create inconsistency; effective service systems define standards and empower appropriate recovery.

10. Illustrative Case: Improving Medication-Order Accuracy

Background. A fictional regional hospital pharmacy processes approximately 4,000 medication orders per week. The target is accurate dispensing and timely delivery to clinical wards.

Problem. Monthly data show rising correction requests. Staff initially attribute the problem to individual carelessness, but the quality manager rejects blame as a starting point and forms a cross-functional team.

Analysis. The team creates a SIPOC map and check sheet. A Pareto analysis shows that most corrections involve look-alike packaging, incomplete electronic instructions and interruptions during verification. A cause-and-effect diagram identifies weaknesses in layout, system prompts, replenishment practice, training and workload balancing. A measurement review confirms that correction requests are recorded consistently.

A control chart for weekly correction proportion shows a sustained increase and one point beyond the upper control limit, indicating that the process is not behaving routinely. The team does not interpret this as proof that one employee caused the change. It investigates a recent product substitution and a software-interface change.

Approach. During DMAIC Define and Measure stages, the team establishes a critical-to-quality requirement: the correct medicine, strength, route and patient must be verified before release. In Analyze, it confirms that the interface change increased selection ambiguity. In Improve, the hospital introduces barcode verification, separates look-alike products, redesigns replenishment labels, protects verification time and adds a clear escalation route for ambiguous orders. Staff are trained and the changes are piloted on one shift.

Outcome. After the pilot, the correction rate falls and remains stable. The control chart is updated only after sufficient post-change data are collected; limits are not repeatedly recalculated to hide deterioration. The hospital also monitors stock-outs, dispensing time, staff workload and near misses so that improvement in one measure does not create risk elsewhere.

Learning points. The case demonstrates the value of process ownership, reliable measurement, root-cause verification, human-factors design, SPC and customer safety. It also illustrates ethical governance: patient safety has priority over throughput, and data must not be manipulated to create favourable performance reports.

11. Implementation Challenges, Ethics and Contemporary Practice

Common implementation barriers include weak leadership commitment, poor data definitions, fear of blame, inadequate statistical skill, conflicting targets, unreliable measurement systems and improvement projects that are never sustained. Training should be role-specific: operators need practical control methods, while managers need capability interpretation, risk management and systems thinking.

Quality data raise privacy and governance concerns. Customer, employee and patient information should be collected lawfully, accessed appropriately and retained only as necessary. Automated inspection and AI-based anomaly detection can identify patterns in images, sensor streams and complaints, but models require representative data, validation, monitoring for drift and human oversight. An algorithm can reproduce historical bias or flag correlation without identifying a controllable cause.

Digital quality systems can connect sensors, electronic records, supplier data and corrective-action workflows. However, more data do not automatically produce better decisions. Organisations should define data ownership, validation rules, cybersecurity controls, audit trails and escalation responsibilities. Automation should support—not replace—professional judgement where safety, fairness or complex context is involved.

Best practice is to begin with a customer-relevant problem, establish a trustworthy baseline, verify the measurement system, select an appropriate method, involve process users, test changes on a controlled scale and sustain gains through standard work and management review. Quality should be integrated with cost, delivery, safety, sustainability and employee well-being rather than optimised as an isolated metric.

Chapter Summary

Quality management has evolved from end-of-process inspection to prevention, process control and organisation-wide learning. TQM integrates customer focus, leadership, employee involvement, process thinking, evidence-based decisions and continual improvement. The cost of quality classifies prevention, appraisal, internal failure and external failure costs, allowing managers to evaluate investments in prevention against the consequences of poor quality.

ISO 9001 provides a requirements-based quality management system built around customer focus, leadership, engagement, process management, improvement, evidence and relationships. The seven basic quality tools help teams collect, display, prioritise and investigate information. SPC distinguishes common-cause from special-cause variation and uses control charts to judge process stability. Capability indices such as C_p and C_pk compare stable process performance with specifications, but require appropriate distributional, measurement and stability assumptions.

Acceptance sampling makes lot decisions from samples and involves producer and consumer risks; it is not a replacement for process control. Six Sigma DMAIC provides a disciplined improvement structure, while service-quality management translates customer expectations into reliable, responsive and measurable service processes. Effective quality management combines statistical discipline with leadership, ethics, human factors and customer understanding.

Key Terms

Acceptance sampling: A sampling-based procedure for accepting or rejecting a lot.

Appraisal cost: Cost of inspecting, testing or auditing conformity.

Capability: The ability of a stable process to meet specification requirements.

Common cause: Routine variation inherent in a process system.

Control chart: A time-ordered chart with a centre line and statistically calculated control limits.

Control limit: A statistical boundary indicating expected process behaviour.

Cost of quality: Prevention, appraisal and failure costs associated with quality.

Critical-to-quality characteristic: A measurable feature important to customer or process requirements.

DMAIC: Define, Measure, Analyze, Improve and Control.

External failure: A defect discovered after delivery to the customer.

Internal failure: A defect discovered before delivery.

ISO 9001: An international standard specifying quality-management-system requirements.

Nonconformity: Failure to meet a specified requirement.

Process capability index: A numerical comparison of process performance with specification limits.

Quality: Degree to which characteristics fulfil requirements and customer needs.

Special cause: An unusual, identifiable source of process variation.

Statistical Process Control: Use of statistical methods to monitor and improve process behaviour.

Total Quality Management: Organisation-wide philosophy of customer-focused continual improvement.

Variation: Differences among process outputs or observations.

Review Questions

5 Short Answer Questions

1. Distinguish common-cause variation from special-cause variation.
2. What is the difference between a control limit and a specification limit?
3. Name the four categories in the cost of quality.
4. State the five stages of DMAIC.
5. Why should a process be stable before capability is assessed?

5 Long Answer Questions

1. Explain the principles of TQM and discuss the role of leadership in implementation.
2. Describe how ISO 9001’s process approach can improve an organisation’s QMS.
3. Explain the seven basic quality tools and give an appropriate use for each.
4. Compare SPC with acceptance sampling, including their purposes and limitations.
5. Discuss how service organisations can measure and improve customer-focused quality.

3 Analytical / Application Questions

1. A process is stable but has a low C_pk because its mean is close to the upper specification limit. Analyse two possible management responses and identify risks associated with each.
2. A Pareto chart shows that one defect category is frequent, while another is rare but creates serious safety consequences. Design a prioritisation approach that does not rely on frequency alone.
3. Develop a DMAIC project outline for reducing late deliveries in a service operation. Specify the customer requirement, measures, likely data sources and control plan.

Multiple Choice Questions with Answers

Question 1

Which activity is primarily a prevention cost?

A. Warranty replacement

B. Operator training

C. Final product inspection

D. Scrapping defective units

Correct Answer: B. Operator training

Question 2

What does a control chart primarily assess?

A. Whether specifications are commercially attractive

B. Whether a process is statistically stable over time

C. Whether every unit is defect-free

D. Whether a supplier has the lowest price

Correct Answer: B. Whether a process is statistically stable over time

Question 3

A process has USL = 50.5, LSL = 49.5 and σ = 0.10. What is Cp?

A. 0.67

B. 1.00

C. 1.67

D. 3.33

Correct Answer: C. 1.67

Question 4

Which Six Sigma stage verifies suspected root causes using data?

A. Define

B. Measure

C. Analyze

D. Control

Correct Answer: C. Analyze

Question 5

Acceptance sampling differs from SPC because acceptance sampling primarily:

A. Makes a lot disposition decision from a sample

B. Eliminates all process variation

C. Sets employee performance targets

D. Calculates customer satisfaction only

Correct Answer: A. Makes a lot disposition decision from a sample

Question 6

Which tool ranks defect categories from largest to smallest and often includes a cumulative percentage line?

A. Histogram

B. Pareto chart

C. Flowchart

D. Scatter diagram

Correct Answer: B. Pareto chart

Question 7

If a process mean moves closer to the upper specification limit while σ remains constant, what generally happens to Cpk?

A. It increases

B. It decreases

C. It becomes equal to Cp automatically

D. It is unaffected in all cases

Correct Answer: B. It decreases

Question 8

Which is an external failure cost?

A. Preventive maintenance

B. Calibration

C. Customer return

D. Process design

Correct Answer: C. Customer return

Question 9

In service quality, reliability most directly means:

A. Providing fashionable premises

B. Performing the promised service accurately

C. Allowing unlimited customer discretion

D. Reducing all employee interaction

Correct Answer: B. Performing the promised service accurately

Question 10

What is the principal purpose of a cause-and-effect diagram?

A. To prove causation mathematically

B. To generate and organise possible causes of a problem

C. To calculate control limits

D. To determine a product’s selling price

Correct Answer: B. To generate and organise possible causes of a problem