Part A: Quality control and assurance
Overview of quality control and assurance
Key takeaways
- Quality control is detective: it inspects, tests and measures output that already exists to confirm that it fulfils requirements.
- Quality assurance is preventive: procedures, competence, calibration, audits and documented evidence, applied before and throughout the work, give confidence that requirements will be fulfilled.
- You cannot inspect quality into a product: inspection finds what the process already made, so a firm relying on it alone pays once to make the defect and again to find it.
- Ishikawa's seven basic tools handle numerical data, while the seven management and planning tools handle ideas, relationships and plans.
- Conformance is fulfilling any requirement; compliance is fulfilling one imposed by an external authority, so a Kenya Standard declared compulsory under the Standards Act (Cap 496) turns non-conformance into an offence.
The distinction between quality control and quality assurance is the single most examined idea in this paper, and it is also the one candidates most often blur. Control detects; assurance prevents. Everything else in the chapter follows from that.
Quality, control and assurance defined
Quality control and quality assurance compared
| Quality control | Quality assurance | |
|---|---|---|
| Purpose | Fulfil requirements | Provide confidence that requirements will be fulfilled |
| Orientation | Detective, reactive | Preventive, proactive |
| Object | The product or output | The process and the system |
| Timing | During and after production | Before and throughout |
| Responsibility | Usually a specific function or the operator | Everyone, coordinated by the system |
| Typical activity | Inspection, testing, measurement, sorting | Procedures, training, calibration, audit, supplier approval |
| Question asked | Is this item conforming? | Is the process capable of producing conforming items? |
| Failure signal | A defective item found | A weakness in the system found before it produces a defect |
| Cost category | Appraisal cost | Prevention cost |
You cannot inspect quality into a product
Deming's point, and the whole justification for assurance. Inspection at the end of a process discovers what the process already produced; it does not change what the process produces. A firm relying on inspection alone pays twice, once to make the defect and once to find it, and still ships whatever the inspection misses.
Control and assurance in practice
Quality control and assurance in practice
| Kenyan setting | Quality control activity | Quality assurance activity |
|---|---|---|
| Dairy processor | Testing each batch of milk for antibiotic residues before pasteurising | Approving and auditing farmer suppliers, training on withdrawal periods, calibrating the test equipment |
| Commercial bank | Checking each loan file against the credit checklist before disbursement | Designing the credit policy, training credit officers, and auditing the process quarterly |
| Construction contractor | Cube testing concrete at 7 and 28 days | Approving the mix design, certifying batching plant, qualifying the testing laboratory to ISO/IEC 17025 |
| County hospital | Verifying the patient identity and the drug at the point of administration | Prescribing protocols, competence assessment of nurses, medication error reporting and analysis |
| Tea factory | Measuring fermentation time and moisture on each batch | Maintaining the withering troughs, calibrating moisture meters, training operators, auditing to KS ISO 9001 |
| Software developer | Testing each release against the acceptance criteria | Code review standards, automated regression suites, and the development procedure itself |
Assurance and quality infrastructure
An organisation's own assurance rests on a national infrastructure it does not control, which is the subject of Introduction to Standardization and of Quality Infrastructure and Legal Framework. The connection is direct and examinable.
- Standardization supplies the requirement: without a Kenya Standard, an organisation must write its own specification and the customer has no independent reference
- Metrology supplies traceability: a measurement is meaningful only if traceable through calibration to national measurement standards held by KEBS, otherwise a test result is an opinion
- Testing supplies the evidence, and a test is credible only from a laboratory whose competence has been assessed
- Accreditation supplies confidence in the assessor: KENAS accredits laboratories to ISO/IEC 17025, inspection bodies to ISO/IEC 17020 and certification bodies to ISO/IEC 17065 and 17021-1
- Certification supplies the attestation the market relies on, including the KEBS Standardisation Mark and Diamond Mark of Quality
- Market surveillance supplies enforcement, without which compliance is a competitive disadvantage borne only by honest producers
- The practical consequence for a Kenyan firm: internal assurance is necessary and not sufficient, because a customer or regulator will not accept the firm's own attestation of its own conformity
Internal quality activities
Internal or organisational factors are those the organisation controls, and they are where quality control and assurance are actually exercised.
- Leadership commitment and the quality policy, which determine whether quality holds under commercial pressure
- Competent people: selection, training, competence verification and awareness under KS ISO 9001 clause 7.2 and 7.3
- Documented procedures and work instructions, so that the method survives the person who devised it
- Capable equipment: process capability, maintenance and calibration
- Controlled inputs: supplier selection, incoming inspection and specification
- In-process control: check frequencies, control charts, first-off approval and line clearance
- Final inspection and release, with defined authority to release or hold
- Control of nonconforming output and of concessions
- Internal audit, corrective action and management review
- The work environment: temperature, humidity, cleanliness, lighting and noise, which for food, pharmaceutical and laboratory work are process parameters rather than amenities
- Information systems that get quality data to the person who can act on it while it is still actionable
External quality activities
External or environmental factors originate outside the organisation and must be managed rather than controlled.
- Customer requirements and specifications, including a supermarket's or an export buyer's own standard, which frequently exceeds the regulatory minimum
- Statutory and regulatory requirements: compulsory Kenya Standards under the Standards Act (Cap 496), sector regulation, NEMA licence conditions, and the Food, Drugs and Chemical Substances Act (Cap 254) for food
- Supplier performance, which the organisation influences through selection, specification, audit and development but does not control
- Third-party certification and surveillance audits
- Market surveillance by KEBS and by sector regulators
- Accredited external testing and calibration
- Consumer feedback, complaints and, increasingly, public commentary on social media
- Competitive benchmarks, since a specification that was adequate becomes inadequate when a competitor exceeds it
- Availability of infrastructure: power quality, water quality, transport conditions and network availability all affect achievable quality and none is within the firm's control
Quality control tools
Ishikawa's seven basic quality tools are the examinable set, and the syllabus expects the candidate to know what each is for rather than merely to name it.
The seven basic quality control tools
| Tool | Purpose | When to reach for it |
|---|---|---|
| Check sheet | Structured collection of data at the point of occurrence | Before any analysis: you cannot analyse data you did not collect systematically |
| Histogram | Show the distribution of a measured characteristic | To see centring, spread and shape, and whether the process fits the tolerance |
| Pareto chart | Rank causes by frequency or cost with a cumulative line | To identify the vital few causes worth attacking first |
| Cause and effect (Ishikawa or fishbone) diagram | Organise possible causes into categories | In a team, to ensure no category of cause is overlooked |
| Scatter diagram | Show the relationship between two variables | To test whether a suspected cause actually correlates with the effect |
| Control chart | Distinguish common-cause from special-cause variation over time | To decide whether a process has changed or is merely varying |
| Stratification (or flow chart, in some lists) | Separate data by source: machine, shift, operator, supplier | Whenever an aggregate figure conceals a difference between sources |
- Deployment sequence in practice: check sheet to collect, stratification to separate, Pareto to prioritise, fishbone to hypothesise causes, scatter to test them, histogram and control chart to characterise the process, and then corrective action
- The tools are deliberately simple so that operators, not only specialists, can use them, which is Ishikawa's central point about company-wide quality control
- Deployment requires three things beyond the tools themselves: training in their use, data that is actually collected, and management willingness to act on what they show
Quality assurance tools
Conformance and compliance
- In Kenya, a Kenya Standard is voluntary until declared compulsory by the Cabinet Secretary under the Standards Act (Cap 496); once declared, non-conformance becomes non-compliance and an offence
- Evidence of conformance is the records required by clause 8.6: evidence of conformity with acceptance criteria, with traceability to the person authorising release
- Evidence of compliance additionally requires whatever the regulator prescribes: permits, marks, declarations, notifications and returns
- A compliance register listing every applicable requirement, its source, its owner and how conformity is evidenced is the practical instrument, and it is what an auditor asks for first
- Non-compliance consequences range from detention of goods and withdrawal of a permit to prosecution and, for company officers, personal liability
Quality management tools
The seven management and planning tools complement the seven basic tools: where the basic tools handle numerical data, these handle ideas, relationships and plans.
The seven quality management and planning tools
| Tool | Use |
|---|---|
| Affinity diagram | Group large numbers of ideas from a team into natural themes |
| Relations (interrelationship) diagram | Map cause and effect among many factors to find the drivers and the outcomes |
| Tree diagram | Break an objective down into progressively more detailed actions |
| Matrix diagram | Show the strength of relationship between two or more sets of items, for example requirements against process steps |
| Matrix data analysis | Analyse a matrix numerically to reveal patterns |
| Arrow diagram (network) | Schedule interdependent activities and identify the critical path (Project Management) |
| Process decision programme chart | Anticipate what could go wrong in a plan and prepare countermeasures |
- Also examinable in this family: quality function deployment (the house of quality), which translates customer requirements into technical characteristics
- Benchmarking, which compares performance and practice against a better performer
- The 5 Whys, for cause chains simple enough not to need a fishbone
- The balanced scorecard, treated in the chapter on performance measurement tools as a performance measurement tool
- Lean tools such as value stream mapping and 5S, and Six Sigma's DMAIC cycle, developed in Quality Improvement and Excellence Models
Answer QC-versus-QA questions with the four-part test
Purpose (fulfil requirements against provide confidence), object (product against process and system), timing (during and after against before and throughout), and orientation (detective against preventive). Give the four contrasts, then one Kenyan example of each, and the marks are collected. Candidates who write only that 'QC checks and QA prevents' get one mark of four.
Self-check
Distinguish quality control from quality assurance and, using a Kenyan bottling plant, give three examples of each. Explain why an organisation relying on quality control alone will have higher total costs. (12 marks)
Name the seven basic quality control tools and explain the sequence in which you would deploy them to investigate a rise in customer complaints. (10 marks)
Show worked answer
QC and QA distinguished and applied (12 marks: 4 distinction, 4 examples, 4 cost).
Quality control is the part of quality management focused on fulfilling quality requirements. It is detective and reactive, it operates on the product or output that already exists, it takes place during and after production, and it asks whether a particular item conforms. Quality assurance is the part focused on providing confidence that requirements will be fulfilled. It is preventive and systemic, it operates on the process and the management system, it takes place before and throughout production, and it asks whether the process is capable of producing conforming items.
Quality control at a bottling plant: weighing five filled bottles every thirty minutes against the declared volume and plotting them on a control chart; testing each batch of treated water for microbiological and chemical parameters before filling; and inspecting cap torque on a sample from each pallet.
Quality assurance at the same plant: calibrating the check-weigher and the torque tester at defined intervals against traceable standards; approving and periodically auditing the preform and closure suppliers, with agreed specifications; and maintaining trained, competence-verified operators together with documented line-clearance and changeover procedures, verified by internal audit.
Why control alone costs more: quality control finds defects, it does not prevent them, so every defect it finds has already consumed materials, machine time and labour. The organisation pays to make the defective item and pays again to detect it, and then pays a third time to scrap or rework it. Because inspection is never fully effective, particularly where it is by sample or by eye, a proportion escapes to the customer, generating complaints, replacement, freight, credit notes and the far larger cost of lost custom. The cost-of-quality argument in Quality Improvement and Excellence Models makes the relationship explicit: prevention costs are the cheapest, appraisal costs are the next, internal failure costs more, and external failure costs most, commonly by an order of magnitude at each stage. Assurance moves expenditure from the expensive categories to the cheap one. An organisation relying on control also never learns why the defects occur, because inspection produces a count of failures and not an explanation, so its defect rate is stable rather than falling, and its inspection cost is permanent rather than reducing.
The seven tools and their sequence (10 marks: 4 for the tools, 6 for the sequence).
The seven basic quality control tools are the check sheet, the histogram, the Pareto chart, the cause and effect (Ishikawa or fishbone) diagram, the scatter diagram, the control chart, and stratification (some lists substitute the flow chart).
Sequence for investigating a rise in complaints:
1. Check sheet first, to collect the complaint data in a structured way: by product, date, batch, defect type and severity. If complaints are recorded only as free text, no analysis is possible, so structuring the collection is the first act and it may itself take a month.
2. Stratification next, to separate the aggregate by line, shift, machine, supplier, batch and period. The commonest single finding in a complaint investigation is that an apparent plant-wide rise is confined to one line or one shift, and stratification is what reveals it. An unstratified average conceals exactly the information needed.
3. Pareto chart, to rank complaint types by frequency and separately by cost, and identify the vital few accounting for most of the problem. Ranking by cost as well as frequency matters, because a rare complaint with a safety dimension outranks a common cosmetic one.
4. Cause and effect diagram, in a team including operators, to generate possible causes of the leading complaint type across categories such as machine, method, material, manpower, measurement and environment, so that no category is overlooked and the investigation does not fix on the first plausible cause.
5. Scatter diagram, to test whether a suspected cause actually relates to the effect: for example plotting complaint rate against a process parameter, a supplier or an ambient condition. This converts hypotheses into evidence and eliminates the ones that do not hold.
6. Histogram, to examine the distribution of the relevant measured characteristic, showing whether the process is centred, how wide its spread is, and whether it fits the tolerance at all, which distinguishes a centring problem from a capability problem.
7. Control chart, to establish over time whether the process has genuinely changed, and if so when, which frequently points directly to what changed on that date: a supplier delivery, a maintenance intervention or a personnel change. It also prevents the error of reacting to common-cause variation.
The output then feeds corrective action under clause 10.2, and effectiveness is verified by continuing the same charts after the change.