Lesson 5: Quality Management / Statistical Quality Control Flashcards
What is Quality?
Meeting or exceeding the expectations of customers
In several fronts such as:
- Conformance to specifications
- Perceived value
- Fitness for use
- Support/services
- Psychological impression
What are the dimensions of quality?
Performance - main characteristics of the product or services.
Aesthetics - appearance, feel, smell, taste.
Special Features - some extra characteristics.
Conformance - how well the product or service conforms to customer’s expectations.
Reliability - consistency of performance.
Durability - useful life of the product.
Perceived quality - indirect evaluation of quality, such as (e.g. reputation).
Serviceability - service after sale.
How is service quality evaluated?
Tangibles - the physical appearance of facility, equipment, personnel, and communication materials.
Convenience - the availability and accessibility of the service.
Reliability - the ability to perform a service dependably, consistently, and accurately for a certain length of time.
Responsiveness - the willingness of service providers to help customers in unusual situations and to deal with problems.
Time - the speed with which service is delivered.
Assurance - the knowledge exhibited by personnel and their ability to convey trust and confidence.
Courtesy - the way customers are treated by employees.
Consistency - the ability to provide the same level of good quality repeatedly.
Benefits of Quality
Quality products fetch premium prices in the market and create a loyal customer base. In addition, costs related to warranty expenses are lowest for such companies.
Improved Quality of Design & Reputation; Premium Prices; Loyal Customer Base; Lower Warranty Expenses.
Quality Improves Perceived Value, without acceptable quality, products/services cannot qualify for order qualifiers.
Costs of Quality
Failure Costs:
Internal Failure Costs - fixing problems during production. (Yield losses, Rework)
External Failure Costs - fixing problems after delivery to customer. (Warranty, Consequential losses (Additional costs), Litigation cost, Reputation/Future revenue)
Appraisal Costs:
Appraisal costs are associated with inspection and testing. The cost comes down as prevention matures.
Helps in identifying the quality problems; Cost incurred in inspection and quality checks at various levels; Quality test development; Screening of defective product.
Prevention Costs:
Prevention costs are associated with preventing the defects before they happen. This may involve quality training, planning, customer assessment, creating SOPs (this is standard operation processes) etc. Prevention increases expenditure of time and money and may delay the product.
Redesigning the process; redesigning the product; training of employees; suppliers involvement.
Cost of detecting a defect
The cost of detecting a defect grows exponentially as the product moves from the conceptual stage to manufacturing, testing and finally to customer.
Identifying the flaws in the process and testing stages may result in delay, but identifying them at the customer stage may bring recalls, lawsuits, bad reputation, etc.
Quality Gurus
W. Edwards Deming gave 14 steps to Quality and emphasized on reducing variations, using statistical process control, and involvement of top management.
Joseph M. Juran focuses on continuous improvement, which includes planning control and improvement. He believes that 80% of quality defects are controllable, thus management has the responsibility to correct these deficiencies.
Armand Feigenbaum introduced the concept of “quality at the source”. For example, defects should not be passed to the next workstation.
Philip B. Crosby emphasize that Quality is free – without Quality you have to pay at the customer level and the cost will be very high. But by improving the design and process, we can save the failure costs. The cost of adopting the good quality practices will cost less than the failure costs and hence quality is free. Moreover, he believed that any level of defects is too much and management should introduce programs to target zero defects.
Certifications
- ISO 9001
- ISO 14000
(Check Formula Sheet for more)
What is Total Quality Management (TQM)?
Total Quality Management is a philosophy that involves everyone in an organization in a continual effort to improve quality and achieve customer satisfaction.
TQM requires the participation of anyone who is related to Quality. For example, suppliers of raw material and store keepers who unload the material and store it.
What is Continuous Improvement (TQM)?
The philosophy of continually seeking ways to improve process (Kaizen).
Ways of doing this are:
- Identify benchmarks of excellent practices
- Continually improve the operation/process
- Develop a sense of operator ownership
- Involve all employees
What is employee involvement (TQM)?
Quality is everyone’s responsibility, which means that everyone who is related to creating or producing services and products should involve themselves in detecting and correcting the quality issues. In other words, all employees should participate and should be given authority to contribute.
Employee empowerment is one of the pivotal ideas common to all quality improvement philosophies.
Employees can participate and provide solutions in many ways. Some of the most common approaches are:
Quality circles: A group of workers in the same department who meet to discuss ways of improving the products or processes.
Brainstorming: brainstorming is a technique for generating a free flow of ideas on finding causes and solutions.
Affinity diagram: An affinity diagram shows the relationships among large number of ideas.
There are also special purpose teams, and self-managing teams.
What is customer satisfaction (TQM)?
The third goal of TQM is customer satisfaction, which means meeting or exceeding customer expectations.
TQM looks at the quality differently.
In a traditional quality management setting, the focus is only to look at the final product or services.
In TQM, we look at every aspect related to the product, from purchasing processes to manufacturing processes and beyond.
TQM Philosophy
In the center, our focus is customer satisfaction – which improves with superior quality and services. Superior quality and services is achieved through continuous innovation and improvements and with the participation of employees who create the products and services. Continuous improvement and employee involvement requires the setting of new goals, using new designs of making tools, designing and improving processes, involving up stream partners which improves the purchasing, etc.
Deming Wheel (PDSA):
Plan: Studying the current problem, document the problem, collect data, specify measures for evaluation and develop the plan.
Do: Implement the plan, document any changes made during this phase, collect data systematically for evaluation.
Study/check: Evaluate the collected data and check how closely results match the original goals of the plan.
Act: If the results are successful, standardize the new solution or technique, document it. If not, revise the plan and repeat the cycle.
Which of the following is not a dimension of service quality?
- Conformance
- Tangibles
- Responsiveness
- Convenience
- Assurance
Conformance
Defective material from suppliers and lost production time are examples of:
- Internal failure costs
- External failure costs
- Replacement costs
- Prevention costs
- Appraisal costs
Internal failure costs
Costs related to inspections, testing, test equipment, and labs are examples of:
- Replacement costs
- Internal failure costs
- Appraisal costs
- External failure costs
- Prevention costs
Appraisal costs
Which of the following is not an example of an external failure cost?
- Scrap and rework during production
- Handling complaints
- Price discounts to offset inferior quality
- Warranty claims
- Loss of customer goodwill
Scrap and rework during production
Monitoring, testing, and correcting quality problems after they occur is known as:
- Perceived quality
- Quality control
- Quality assurance
- Continuous improvement
- Conformance
Quality control
The process of identifying other organizations that are best at some facet of your operations, and then modeling your organization after them is known as:
- Industrial espionage
- Employee empowerment
- Continuous improvement
- Parody
- Benchmarking
Benchmarking
Groups of workers who meet informally to discuss ways to improve products or processes are called:
- Quality circles
- Quality teams
- Continuous improvement teams
- Benchmarking teams
- Brainstorming teams
Quality circles
True or False:
Total quality management explicitly recognizes that management is primarily responsible for quality, not the workers with direct responsibility for completing work tasks.
False
True or False:
A company that commits to TQM adopts a process-oriented focus rather than the product-oriented focus which traditional organizations typically have.
True
True or False:
Cost of inspectors, testing, test equipment, and labs are examples of appraisal costs.
True
True or False:
Broadly defined, quality refers to the ability of a product or service to consistently meet or exceed customer expectations.
True
Common Quality Tools
Checklists; Histograms/Bar Charts; Pareto Charts; Scatter Diagrams; Cause-and-effect Diagrams; Graphs; Control Charts.
Checklists
Checklist are for data collection. To record frequency of occurrence of certain characteristic of service or product.
Histograms/Bar Charts
In histograms and bar charts, we plot the frequency of desired information. It is useful for a comparative study. For example, we can plot number of defective products produced either every hour or every shift or machine etc.
Similarly, we can plot average customer waiting time with respect to every customer service representative, every day of the week, etc.
Pareto charts
Pareto charts help answer the question of which quality problem should be addressed first? Named after Vilfredo Pareto, these charts tend to show the 80-20 rule: that 80% of the activity is caused by 20% of the factors.
Scatter diagrams
Scatter diagrams are useful in finding the correlation between two variables. For example, is there any relation between fatigue (continuous working hours) and number of defective products?
Another correlation example is that the motor of newly produced blenders produces excessive vibration with rise of motor temperature.
Cause-and-effect
Cause-and-effect diagrams are also known as fish-bone diagram. It helps us in finding the cause and effects.
For example, if we see a low number of students show up to class in one given day, we can identify several causes.
It could be due to transport problem, or some big public event or due to exam preparation, etc. Once we have effects, we can identify the causes.
Graphs
Graphs or run charts are very similar to scatter diagrams or bar charts. We plot data points with respect to time to observe any trends. In fact, run charts are produced continuously and based on the observation. Other tools are used to investigate further.
Control charts
Control charts are used to plot some attributes or dimensions of the products – very similar to graphs. However, in control charts, we also plot control limits and we observe if each data point is within the limits.
Benchmarking
Benchmarking is one way of improving quality performance. It involves planning, analysis, integration and action.
Planning involves identifying the process in your organization that needs an improvement, identifying a firm for bench marking, identifying performance measure and then data collection.
Data collection will help in establishing the gap and identifying the causes of low performance.
Integration involves setting up the goals and support, and in action, we develop a plan to improve the process, implement the improved process, monitor the progress, etc.
Quality Engineering
An approach that involves combining engineering and statistical methods to reduce costs and improve quality by optimizing product design and manufacturing process.
Quality Loss Function (Taguchi Loss Function)
Deviation of Quality characteristics from the target values.
Taguchi was a pioneer in quantifying the quality in the form of function – mostly known as Quality loss function.
Taguchi Concepts include: Quality robustness, Quality loss function, and Target-oriented quality.
Six Sigma
Sigma is a Greek notation, commonly used to represent the standard deviation. Higher standard deviations mean there is more variability around the mean and hence, a higher chance of unacceptable Quality.
Typically, the central area of normal curves can be represented in terms of sigma.
It simply says that if the deviation in the process is so small, that is, sigma is so small, then the product’s specification limits can cover roughly all of the normal curve and we may not be making any defective products. For instance, 6 sigma quality level means we may be producing 3.4 defects per million, which means, whatever the variability we are producing, it is still within the specification limits.
Quality Tools Problems Solve %
90% of the quality problems can be fixed by basic quality tools and philosophy.
Improving the process further requires the Six Sigma quality philosophy.
90% - Basic tools of quality
< 10% - Six Sigma
< 1% Outside Specialists
Six Sigma Improvement Model
Define - Measure - Analyze - Improve - Control
Define (Six Sigma Improvement Model)
Determine the characteristics of the processes which are critical to customer satisfaction and identify any characteristics and the process capabilities.
Measure (Six Sigma Improvement Model)
Quantify the work the process does that affect the gap. Decide what to measure, Identify data source, Data collection.
Analyze (Six Sigma Improvement Model)
Perform process analysis—process improvement and redesign. Includes statistical process control, cause-effect diagram, scatter diagram, etc.
Improve (Six Sigma Improvement Model)
Modify or redesign existing methods to meet the new performance objectives. Implement the changes.
Control (Six Sigma Improvement Model)
Monitor the process to make sure that high performance levels are maintained. Again, data analysis, etc., will help with monitoring and evaluation of relative progress.
A fishbone diagram would be used to:
Categorize potential causes of a problem.
A quality improvement technique that involves the sharing of thoughts and ideas in a way that encourages unrestrained collective thinking is:
Brainstorming
A chart showing the number of occurrences by category would be used in:
A Pareto analysis
The tool that is useful to investigate the relationship between two variables is:
A scatter diagram
True or False:
Taguchi quality loss function is a graphical representation of how an increase in deviation from the target value leads to a faster rate of increase in customer dissatisfaction.
True
True or False:
The purpose of benchmarking is to establish a standard against which the organization’s performance can be judged, and to identify a model for possible improvement.
True
True or False:
Organizations committed to Six Sigma programs have very capable and precise processes.
True
Statistical Quality Control
Uses statistical techniques and sampling to monitor and test the quality of goods and services, in order to avoid 100% inspection.
Acceptance Sampling
Relies on the inspection of a sample from a batch to determine whether the batch is acceptable or not.
Statistical Process Control
Determines if the process is operating within the acceptable limits during the production.
Double Sample Plan
If the first sample failed the inspection, then we pick another sample.
If the combined defectives in both samples are less than a predetermined number (c2), we keep the lot. Otherwise, we reject it.
Sampling Errors
Since, we are accepting the lot based on the inspection of a very small number of items, our approach could give the wrong conclusion. We may pick a defective lot or we may reject a good lot based on our sampling.
These errors are commonly known as type 1 and type 2 errors.
Type 1 Error
When the part is good and our experiment says it is bad. This increases the probability of rejecting good parts as scrap.
Type 2 Error
When part is bad but we accept it as a good. This increases the probability of a customer getting a bad lot represented as good.
Sampling Plan Careful Design
Overall expected cost due to sampling (which includes costs of accepting defective items) is smaller than the costs of inspecting more or all items.
It achieves the purpose of sampling- saving time, space, capital.
To ensure the acceptable quality level (maximum percentage of defects that a company is willing to accept).
True or False:
If a lot, or batch, is “rejected” during acceptance sampling, the lot will usually be thrown away.
False
True or False:
In a single-sampling plan, the entire lot, or batch of items, is accepted or rejected based upon only one specified sized sample.
True
True or False:
Acceptance sampling is most useful when the cost consequences of passing defectives are low.
True
True or False:
Acceptance sampling plans must specify the lot size, the sample size, and the acceptance/rejection criteria.
True
True or False:
In a double-sampling plan, a second sample is taken if the results of the first sample are inconclusive.
True
Acceptance sampling plans might call for selection of all of the following except:
- 100% inspection plan
- A single sample
- Two samples
- Multiple samples
- Several samples
100% inspection plan
A Type II (beta) error occurs when:
A bad lot is accepted
A Type I (alpha) error occurs when:
A good lot is rejected
Process Control Charts
Statistical techniques used to identify these variations and label them as random or assignable.
Introduced by Walter A. Shewhart (1920), Bell Laboratories
Around 100 years in existence
Used as a tool to assist in controlling variation
Japanese manufacturing became more competitive by using these tools
Assignable Variations Identified
The process may not be capable and our intervention is required in order to investigate the cause of unnatural variations.
Random variation
Natural variations in the output of process, created by countless minor factors.
Assignable variation
A variation whose source can be identified
Main task of SPC (Statistical Process Control)
Distinguish assignable from random variation
Normal Distribution (Quick Info)
If some samples are outside the 3 sigma limit, the probability of such an event happening is very small or very rare. We conclude that it may not be due to random variation and an investigation is necessary.
Type 1 and 2 Errors SPC
1: Concluding a process has changed when it has not
2: Concluding a process is in control when it is not
Control Charts for Variables
Sample Mean Chart; Sample Range Chart;
Sample Mean Chart
Sample mean control charts are popularly known as X bar chart. In this chart, we track our sample means and if the sample means deviate much i.e. falls outside the control limits, we mark the process out of control.
Sample Range Chart
Sample range chart are popularly known as R charts. In this chart, we compute the range of a given sample and plot the range.
These expressions provide the control limits for the Range chart.
D3 and D4 values are provided in the same table.
Typically, R charts always plot in association with X-bar charts because X chart tries to monitor the Mean of the process and R monitors dispersion of the sample. If any of the chart is out of control, we label the process out-of-control.
Control Charts for Attributes
Control charts for attributes are used when the process characteristic is counted rather than measured.
P-charts; C-charts.
P-charts
P stands for proportion or probability.
So, if the items can be easily categorized as good or bad, then it will be very easy to compute proportion defectives in a lot or compute the probability of defective items in a sample/lot.
In p-charts, we pick samples and in each sample we find the number of defective items.
The number of defective items divided by the sample size will give us the proportion of defectives or percentage of defectives.
Based on that, we compute the control limits.
The p-chart is used to monitor the fraction of defectives in samples.
These expressions are used to compute the control limits. sigma limits are commonly used so we use z=3
Note: Because the formula for control limits is an approximation, sometimes LCL will be negative. In this case, zero should be used as the lower control limit.
C-charts
The c-chart is used to monitor the number of defects per unit.
These expressions are used to compute the control limits. 3 sigma limits are commonly used so we use z=3.
If the value of c is unknown, as is generally the case, the sample estimate is used in place of c. When the lower control limit is negative, it is set to zero.
C-charts are used when the number of occurrences of some attributes per unit is measured. Such as: Scratches, chips, dents, or errors per item; Cracks or faults per unit of distance; Breaks or tears per unit of area.
Non-Random Patterns in Control Charts
In control charts, we see if the variations are randomly distributed around the central line. If they are not, then they must be forming some pattern – for example, three or more points on one side of the central line, or showing upward or down ward trend, etc.
Remember, we say that if a point is outside the control limit, then the process needs attention – because the probability that a normal process shows such behavior is very small – or we call it a rare event.
The control chart is used to identify such rare events. In other words, if the process is normal and points are forming some patterns, the probability of such event is very small and thus if such is happening, then the process requires attention.
This chart shows commonly non-random patterns. In other words, even when all the points are within the control limits, these processes require attention.
True or False:
Quality control efforts that occur during production are referred to as statistical process control.
True
True or False:
The optimum level of inspection minimizes the sum of inspection costs and the cost of passing defectives.
True
True or False:
Concluding that a process has not changed when it has is known as a Type I error.
False
True or False:
Concluding that a process has changed when it has is known as a Type II error.
False
True or False:
Statistical process control is based on comparing periodic samples from a process to predetermined limits.
True
True or False:
Assignable variation is variation due to a specific cause, such as tool wear.
True
True or False:
A sample statistic that falls outside the control limits suggests that the process mean has changed.
True
True or False:
Sample range control charts are used to monitor process dispersion.
True
True or False:
If a point on a control chart falls outside one of the control limits, this suggests that the process output is non-random, assignable variation.
True
True or False:
A p-chart is used to monitor the fraction of defectives in the output of a process.
True
True or False:
A c-chart is used to monitor the number of defects per unit of a process output.
True
True or False:
The number of defective parts in a sample is a process characteristic that is counted rather than measured.
True
True or False:
A process is “capable” if the process output falls within the design specification.
True
The purpose of control charts is to:
Distinguish between random variation and assignable variation in a process
As the sample size increases, the distribution of sample averages approaches a Normal distribution regardless of the shape of the sampled population. This is identified in which of the following?
- Assignable Variation
- Random Variation
- Central limit theorem
- Type I errors
- Mean charts
Central limit theorem
A control chart used to monitor the process mean is the:
x-bar chart
Using three sigma control limits (rather than two sigma limits) tends to have what impact on error probabilities?
Increase the Type II and decrease the Type I.
A control chart used to monitor the fraction of defectives generated by a process is the:
p-chart
The range chart (R-chart) is most likely to detect a change in:
Process variability
A c-chart is used for:
Number of defects per unit
You want to prepare an R-chart. If the number of observations in a sample is 5, what is the appropriate “factor” used in the computation of the LCL?
The factor used in the computation of the UCL and LCL is 0 (from Table 10-3).
You want to prepare a p-chart and you observe 200 samples with 10 in each and find 5 defective units. What is the resulting “proportion defective”?
The resulting proportion defective is 0.0025 since 5/(200x10) = 0.0025.
You want to prepare an x-bar chart. If the number of observations in a “subgroup” is 10, what is the appropriate “factor” used in the computation of the UCL and LCL?
The factor used in the computation of the UCL and LCL is 0.31 (from Table 10-3).
Process Capability
Ability of the process to meet the design specifications for a service or product.
Capability Analysis
The determination of whether the variability of the process output falls within the acceptable range of variability allowed by the design specification.
Capable (Capability Analysis)
Process output falls within specifications
Incapable (Capability Analysis)
- Redesign process or reduce variability
- Use alternative process
- Use 100% inspection
- Examine/relax design specification
Cookie Packages (Cpk)
An index that shows how well the units being produced fit within the specification limits.
The process is considered capable if Cpk greater or equal to 1.
This process will produce a relatively high number of defects.
Many companies look for a Cpk of 1.3 or better. Six Sigma companies want 2.0.
A methodology that is used to show how well parts being produced fit into a range specified by design limits is:
Capability Analysis
True or False:
Control limits and process variability are directly related.
True
True or False:
“Process capability” compares “process variability” to the “design specifications”.
True
True or False:
The process capability index, indicated by Cp is calculated as the ratio of the design specification width to the process width.
True
True or False:
The process capability index, indicated by Cpk is used only when the process is centred.
False
True or False:
In order for a process to be capable, it must have a capability ratio of at least 1.00.
True
