Philip Crosby, one of the forefathers of modern-day quality management, titled his seminal work on quality management, “Quality is free.” When we say that quality is free, we mean to say that quality shouldn’t be thought of as an additional expense. If we are willing to make that initial commitment to quality from day one, we will see the returns immediately.
There is a significant investment to be made from day one, but this is not necessarily measured in monetary terms. However, once that investment is made, then there are processes in place for quality and people generally have a mindset to do extremely good work and satisfy the requirements set forth by the client, whether that be an internal or external client.
Quality Concepts covered in this section
There are several Quality management concepts that we will cover in this section
- Quality Management
- Quality Planning
- Quality Assurance
- Quality Control
- Quality Control Tools
- Continuous Improvement
- Just in time
- Impact of Poor Quality
- Cost of Quality
PMI’s Project Quality Management knowledge area is firmly grounded in several well-known and commonly adapted philosophies relating to quality. We will cover these concepts in this section.
The concept of zero defects is that there is no deviation that is acceptable when it comes to quality. In other words, we cannot deviate from our requirements or specifications. According to Crosby, “Quality is Conformance to the Requirements and Specifications.”
The key term to look for in the exam is ‘Zero Deviations’. It is not a slogan, but rather a standard to aspire to. The questions might want to distract you by offering options such as “goal” or “objective”. Zero Defects is neither a goal nor an objective, but rather a standard by which we live. We will not accept anything less when it comes to quality from Crosby’s point of view.
We must also understand that we should not give the customer more than they expect. If you give someone more than they asked for, you are essentially ‘Gold-plating’ and that is wasteful in terms of time and money and more importantly, the customer is not expecting it. If you define your requirements correctly and accurately, then the customer should be 100% satisfied when you give them exactly what they ask for and nothing more.
Fitness for Use
One additional factor to consider when evaluating quality is fitness for use. We must consider whether the product or service is fit for the use for which it was originally intended.
Prior to W.E. Deming, it used to be thought that all processes had defects and the only way to eliminate the defects in a finished product was to hire quality inspectors to stand by a production line and look for these defects. Known as quality by inspection, this antiquated approach towards quality management was in use for many years and ultimately found to be extremely expensive, since there was no way to realistically inspect every item that came off the production line.
Conventional thinking has fundamentally changed how we approach quality. We now believe that defects can be avoided and we can reach zero defects for any process. The major focus here is that we plan for quality at the inception of a project in rather than try to work quality in to the project later on. Quality should be planned into the project rather than being inspected in.
The concept of Quality Assurance can be thought of as a holistic, company-wide approach towards quality. We tend to see large-scale quality assurance programs in organizations that permeate through all of the levels of hierarchy within a business, including the project and its supporting elements within the organization. Quality assurance when done properly typically starts off as a managerial initiative. It is important for management to build up a culture that chooses to focus on quality in all things that the company does. We call this a top-down approach towards quality.
Audits and Evaluations
You need to be able to distinguish between audits and evaluations in the exam. What kind of quality audits are you going to use? It includes all of our resources or processes that we bring together that we use to ensure that we produce a quality product.
- Formative Quality Evaluation: This happens during the Project Lifecycle. The evaluation is going on as the project is progressing or forming.
- Summative Quality Evaluation: This happens at the end of a project. This can also be known as lessons learned for all intents and purposes. Our goals are to look at how we functioned during the project; and also to evaluate what processes we used and whether they were effective. We also look at the product or service that we created and we examine how close we were able to get to our quality requirements.
Even though the summative quality evaluation is carried out after the project ends, there should be little or no wait time before the summative evaluation is carried out. This is because relevant information would lose its significance to team members over time. We want to document all experiences while they are fresh in the minds of our team members.
It is also better to meet after each phase or key milestone in a project in order to take a look at what has been completed to date and whether the objectives for that particular part of the project have been achieved.
Responsibility for Quality
When it comes to taking responsibility for quality, we should be aware that responsibility varies depending on the scale or level of activity being performed.
- Q: At the task level, who has the responsibility for quality?
- A: The team member doing the task has the responsibility for quality.
- Q: When it comes to the project, who has the responsibility to quality at the project level?
- A: The project manager now has the responsibility for ensuring quality across the entire project.
The candidate has to be very careful when it comes to tackling such questions in the exam. A subtle change in the question can result in two very different answers.
Cost of Quality
There is a way to measure the cost of attaining and maintaining a certain level of quality for any organization. We can look at this Cost of Quality from a number of dimensions:
- Cost of Conformance: The Cost of Conformance is a proactive means to ensure that a quality product or service is produced by conforming to quality standards. It involves such activities as planning; training and process control. Other costs such as process validation and even the choice of how testing is to be performed are important factors to consider.
- Cost of Non-Conformance: The Cost of Non-Conformance is cost of failure. For example we have rework and repair for a product that did not meet a customer’s level of acceptance. Additional factors such as complaint handling or the damage to reputation can also be considered under the cost of non-conformance. Product recalls are yet another example.
- Internal Costs: The internal cost is the cost of repairing work that you have found yourself.
- External Costs: The external cost is the cost of repair after your client finds the defects. A product recall is a good example of an external cost.
Real or True Cost of Quality
The true Cost of Quality is the cost of Non-Conformance. When asked to pick between the cost of conformance or the cost of non-conformance as the real cost of quality, the answer should be the cost of non-conformance.
Responsibility for the Costs
The Project Manager or the Organization that is performing the work is responsible for paying for the costs of quality. The costs of quality ultimately fall as a responsibility of management. Deming wrote that 85% of the cost of quality was a direct responsibility of management; he later increased that percentage to 90% prior to his death. PMI likes to use figures or statistics in the exam, because that is a great opportunity for them to insert distractor answers. 85% or 90% of the cost of quality is the direct responsibility of our management.
PMI describes quality control as a technical function and not a managerial function. It involves first establishing a technical baseline for the project and then collecting specific data to measure conformance to that baseline.
The quality baseline entails all of the specifications and requirements of what it is that we are going to do.
For example, we might have a list of specifications or requirements or tolerances. This encompasses everything that describes to the greatest detail what the product or service should look like.
Nature of Variables, Attributes and Probability
The following terms are frequently used in relation to experiments and statistical analysis, all relating to Quality Control activities.
- Variable: An example of a variable is a person’s weight can fluctuate over time. Every morning when I step on my bathroom scale, I notice that my weight can go up or down on a daily basis. We can plot those variables and watch my weight go up and down and up and down over time.
- Attribute: Is there a way to turn a variable into an attribute? Yes. The number of times that a person has weighed more than 180 pounds. The variable known as weight now turns into an attribute. Where it was seen previously as a variable, it now is either less than or greater than a specific figure. Your attribute now can take on one of two states, either on or off.
- Probability: For example what are the odds of getting heads or tails when I flip a coin? When I flip a coin three times and I get heads each time, on the forth flip of the coin, what are the odds of getting heads again? My odds are unchanged at fifty-fifty. Because each flip of the coin is an independent event. The outcome of one event does not affect the outcome of a subsequent event. The rule here is not to string together a number of probabilities because they happen to be laid together in the exam.
Relationship between Probability and Distribution
Probabilities are commonly related to some sort of distribution. We refer to a commonly used example where a number of students would bring in candy bars to class, in particular Hershey almond bars. Sometimes when you eat a Hershey Almond bar, it is full of almonds; at other times you find that the entire bar might only have a handful of almonds. This is what we call a distribution. When Hershey bars are made, if you were to randomly sample some of the bars and count the number of almonds in them, you would find that the number of almonds in each bar would almost certainly vary. If you were to count a very large number of candy bars and plot the results, you would get a bell-shaped curve. This is called a normal distribution.
An experiment was conducted where students coming to class would count the almonds in their candy bars and plot the number of almonds found in a bar over several months. The students noticed that during the spring months and summer months, the Almond bars had a very clean, even distribution. There were generally anywhere between thirteen and fifteen almonds per Hershey bar. When it got close to Halloween, they started to see a much larger variation in the number of almonds found in each Hershey bar.
The reason was that the Hershey’s chocolate factory had increased its production levels to anticipate increased demand during Halloween. Due the increased quantity of bars produced, the Hershey’s production lines were not as fastidious with their quality. Sometimes students would find as few as six or seven Almonds per bar and sometimes they would find as many as eighteen or nineteen Almonds per bar.
So the distribution became much broader and flatter as the manufacturer was going through much higher rate of production.
You are required to become familiar with several statistical terms and concepts.
- Mode: The one number that comes up the most often
- Mean: The number that is the average for that distribution. (14 Almonds)
- Median: The middle value in a distribution, of which above and below lie an equal number of values.
Bell Curve Shapes
A tall thin bell curve represents a process that is under control, because all of our samples or population are very close to our mean. This is a good thing. A long flat bell curve represents a process that is under less control, because we have much greater deviation from our mean.
Standard Deviation (S.D.) describes how much of your population you have captured within the diagram and how far they fall from the mean. When it comes to standard deviation, you need to memorize the percentages for the various standard deviations from the mean, depicted by the term sigma (σ). Candidates also should be aware of normal and skewed standard deviations.
1 S.D. : 68%
2 S.D. : 95%
3 S.D. : 99.7%
4 S.D. : 99.99%
5 S.D. : 99.9999%
6 S.D. : 99.999999%
As we look at one standard deviation from the mean, we capture 68% of all of the candy bars. As we move three standard deviations from the mean, we capture a vast majority of the candy bars at 99.73%. This represents most of the Candy Bars manufactured. When we say that a process follows 6σ, we’re indicating that over 99.999999% of our population (i.e. production output) falls within our quality specifications.
The Control Chart gives us our “current capability”. We are measuring our process and we are taking those measurements periodically throughout the project. The measurements serve to tell us how well our process is performing in accordance with our specifications. This is an indicator of how reliable and predictable the process that we established is working.
You might see the following term in the exams. ‘The Voice’ of the process is defined as what your process is telling you it can do. The voice of the customer represents the specifications. The voice of the process determines what we are doing to meet those specifications.
When you have a process that produces a huge number of outputs, it’s sometimes impractical to be able to inspect the population, in other words every single item that is produced. A sample is a smaller subset of your entire population, but is large enough that you feel confident that the results of testing the sample will be the same as testing the entire population.
If you are told you have a valid sample. That is just as good as doing a 100% inspection. A lot of people are uncomfortable with that. They would rather do a 100% inspection. However, if it is a valid sample, it is just as good as a 100% sample.
- Attribute Sampling: The questions you might ask when performing attribute sampling are: Is it on or off, is it heads or tails, and is it one way or the other? Is it above or below 200?
- Variable Sampling: The questions you might ask when performing variable sampling are: Does it fall above or below the range? Where does it fall within the range or general curve and how does that compare to our general population.
- Tolerance: The result in variable sampling that will be acceptable if it falls within a particular range specified. This is called tolerance and tolerance relates to variable sampling.
Quality Control Tools
There are seven tools that are frequently used for Quality Control. You must be familiar with how these tools are used and deployed.
Flowcharts and Diagrams
Questions on flowcharts are fairly easy to answer in the exam. The approach taken by PMI is to ask what advantages there are for each tool and you would use one over the other, etc. You are not required to construct these flowcharts.
We use flowcharts to better understand relationships in a process. A flowchart is a schematic or picture that serves to create a common language or common understanding about a specific sequence. A flowchart looks at a number of events in a sequence and there a number of different types of flowcharts listed in the PMBOK
- Top-Down Flowchart: This represents the management perspective on reality. Life is always moving forward, there is no turning back. There aren’t any looping processes. Fundamentally, a top-down flowchart moves in one direction and it does not allow for looping.
- Classic Detail Flowchart: This is a very valuable and widely used tool in the re-engineering business nowadays. A detail flowchart provides very specific information on a process. In a detail flowchart we have every decision point and every feedback route and every process step. For the exams, know that for the detail flowchart, the box represents the step and a diamond represents a decision point and coming out of the diamond there should be a yes or a no arrow leading down to other elements such as another step or decision.
This is a graphical representation of how work flows through a physical space or facility. Imagine an automobile assembly plant and picture that assembly plant in the form of a graphic with individual workstations and arrows indicating the direction to where the products are flowing. It is just the movement of stuff through physical space. So you can do a workflow diagram on a highway or on just about anything. You can think of how the work moves from point A to point B and what is involved in the process.
From a quality perspective, the work-flow diagram is good for analyzing the flow processes and for planning process flow improvements. This tends to be the angle that PMI takes regarding workflow diagrams in the exam. You may have a scenario which is presented to you and then you are asked which type of flowchart or diagram you should use to analyze this particular scenario. Knowing the value and objective of each of these will lead you to the correct objective. The Ergonomics issue stems from the work-flow diagrams. It deals with ergonomics or human motion in physical space. There are specialized work-flow diagrams associated with particular processes.
Just In Time allows for the movement and storage for small amounts of inventory in a physical space. The entire purpose of the Just-in-time concept is that you do not want to have too much inventory on hand since excess inventory freezes up cash and costs money to store. So you want to build up only as much enough inventory as you consume.
The kan-ban area or system prevents over-production by allowing work to move forward only when the next work area is ready to receive. An actual physical area on the assembly-line or production floor where people have an understanding that this work is coming down but it is not going to come down if I am not ready to receive it. KAN-BAN relates to just-in-time inventory.
Pareto Diagrams (80/20 rule)
Vilfredo Pareto was an Italian economist who said that 80% of the wealth rests with 20% of the population. His concept of 80 versus 20 has been applied generally to almost anything, including quality management. Dr. Joseph Duran took Pareto’s idea and formulated the Pareto principal for quality and he called it the law of the vital few. This is the 80-20 rule and it states that we are going to have 80% of our problems coming from 20% of our items or processes. For anyone who has worked with an inventory application the movement of the inventory you will find that 80% of the activity in an inventory system will be in 20% of the portion of that inventory.
The central concept to the 80/20 rule is that if you can change that one big problem, you are going to have the biggest influence overall on how your process looks and feels in the long term. There are almost limitless numbers of examples when it comes to a Pareto chart. The key to the Pareto chart is which bar you want to fix when it comes to the Pareto chart. Imagine if you will a chart in front of you with a number of vertical bars, one is taller than all of the rest. Which one do you want to fix? It is always the tall one. Do not let any way the question is worded dissuade you from the idea that it always the tall bar which represents the lion’s share of your concerns. That is the one that you really have to go in there and solve in a Pareto chart.
Ishikawa Diagram/ Fishbone Diagram/ Cause & Effect diagram
This tool comes in a variety of names and widely used. It is named very aptly the fishbone diagram because it looks just like the skeleton of a fish. Fishbone diagrams are useful for brainstorming, for examining processes and for sequencing activities.
The cause and effect diagram basically drills down for causes, causes and more causes all driving toward one single, focused effect. We’re always looking for causes and breaking things down into further causes. What is the cause for the heat in the room, it may be some machinery. What is the cause for the machinery, it may be something and what is the cause of that something and we keep going down until you find the entire minutia that can be possibly associated with the effect at the very root of your fishbone diagram.
It is great for brainstorming and looking at problems that you have with quality but it can also be used in a prospective mode, here is an effect that you want to have happen and so what causes do we have to put in place in order to reach that effect. So in quality it can be “let’s see what has occurred and why” or it can be “we want something to occur and what do we need to do in order to make it happen.
The PMP exam will test you on some basic graph styles such as pie charts; line graphs or bar charts. If you see a pie chart and ¾ of it is one thing and you are asked which one of the elements in the chart should you be dealing with first, the answer is simply the large one. Graphs are very powerful forms of communication. PMI says that they should be used liberally in the workplace. Graphs are very helpful for planning improvement processes. Instead of trying to read tables of data, it is much easier to interpret data in graphical form and therefore graphs can become a very popular tool in quality.
This tool may be difficult for some candidates because they’ve had little experience dealing with control charts. There are a few fundamental terms that you need to be aware of:
- UCL – Upper Control Limit
- LCL – Lower Control Limit
To construct a control chart, we take readings of a number of items in our sample and plot them on the chart. If the items fall within the Upper and Lower control limits, they have met our quality standards. In some cases, several patterns emerge that might give us cause to investigate our process. For example, if we take a number of readings and after plotting the data points, we see that these points lie outside of our control limits. These data points are out of control. But is the entire process out of control? Well, we might have cause for investigation.
Rule of Seven:
Let’s say that you have a control chart and you plot seven data points in a row. All seven data points are recorded above the mean. Does that mean a: our process is in perfect control; b: our process is out of control c: there is cause for investigation; d: the process is functioning normally? The answer is that we have cause for investigation. This is because the probability of seven consecutive data points all falling on one side of the mean are extremely low. Imagine flipping a coin and getting heads seven times in a row. Something just doesn’t feel right. For a normal distribution, you would expect to get readings both above and below the mean, since the mean is statisically at the middle of your readings. Seven points on one side is not normal and this should cause us to stop for a minute and take a look at what’s going on.
We also want to know about upper and lower control limits is where the customer’s tolerance limits fits in the whole process. We need to know that the customer’s tolerance fits outside the upper and lower control limits. For example, next time you go into a fast-food place, you might see that they have posted a notice of the mean temperature of what they expect the temperature of their burgers to be. Let’s say that the temperature 185 degrees. Now their system or process will enable burgers to be stored between 180 and 190 degrees consistently. They will always produce burgers within that range. Now you as the customer, if your tolerance limits at that you want a burger between 184 and 186 degrees, then the process may sometimes not serve you. But if you always want your burgers between 175 and 195 degrees, then it is the range you deem acceptable every single time then their process is going to deliver a process that you will eat.
These are the basics of control limits and tolerance limits. The point to remember is that tolerance limits are outside the control limits.
There is a practical question that we want to ask ourselves when designing a process. Can we expect our outputs to be identical in nature no matter how long we keep doing the process? Our outputs will all vary slightly because there is something called natural variation that says it is impossible to make any product with absolute consistency. This is why we have our upper control limits and our lower control limits and understand where our tolerances are. As long as we are within this range, we can have variation and still conform to requirements.
Relationship between sample size and control limits.
If we are doing the assessment of our process on say a hundred units produced, that is going to give us a sample size one set of numbers. What if we do ten thousand numbers? Is it going to change the control limits at all? It is quite possible that the control limits will change, but the good thing is that the larger the sample size, the greater the degree of control we have on that particular process. The larger the process is or the larger the sample size, then the tighter the controls get. This is because the more we are looking at our total population there are fewer outliers in that far end of the process.
You might have gotten one of these the last time you went to rent a car. The pilot and copilot in a commercial air-liner also make frequent use of a checklist before they taxi off onto the runway.
A check sheet differs from a checklist in that it looks at physical flaws. It is looking at specific things that could have gone awry and that a checklist is looking at the specifics of what are the tasks that have to be performed or done. You have to be able to distinguish between the two for the exam. The check sheet also has the name of a measles sheet or measles diagram.
Kaizen – Continuous Improvement
Kaizen is the Japanese word for continuous improvement. Improving quality is not a discrete, one-time event. It happens every time we perform a service or we create a product. We want to do everything incrementally better. It is the small incremental steps, the little things that we can do along the way that can contribute to significant progress.
Sometimes to improve our Organization, we look at other Organizations. We look at the big companies, or the people that have been doing it well. This involves comparing your practices to the practices of other projects or other companies for the purposes of improvements
- Internal benchmark: this is an activity internal to a company, where one branch or department compares its processes to another
- Competitive benchmark: you are comparing your processes to your toughest competitor.
- Functional benchmark: comparing a similar process to your non-competitors.
The human aspect of quality
People need to get really involved about quality. When a person gets highly motivated, then they do a better job in terms of quality. For a time, Ford motorcar company’s primary motto was ‘quality is job one’
When it comes to project management, we have to look at quality against the cost and schedule. We need to ask ourselves, which is more important, quality, cost or schedule? This is very difficult for many people to answer in the exam because many candidates might have taken quality training prior to the exam where they would be introduced to quality with a certain philosophy. Now, that philosophy would be different than what they are used to. According to PMI, modern thinking should emphasize that quality should share equal priority with cost and schedule as it applies to project management. The 3 sides of the triple constraint are time, cost and if you want to look at your requirements (scope) as quality, all bear equal weight.
The impact of poor quality
We need to be prepared for poor quality. What are some of the effects of poor quality? People get less productive, less excited about their work, we increase our costs, there is non-conformance. We wind up spending a lot more money, time and energy on monitoring and evaluation. This is a big problem.
R&M (Reliability / Maintainability)
This is a special category of quality and they are expressed and you will see this in the exam with 2 acronyms, which are MTBF (Mean time between failures) and MTTR (Mean Time To Repair). We’re attempting to measure the reliability by using MTBF which literally asks how long we are going to go before this thing breaks down and when it eventually does break down, MTTR will ask how long it will take before we can bring it back up and running.
Summary: Project Quality Management.
- Quality Management
- Quality Planning
- Quality Assurance
- Quality Control
- Quality Control Tools
- Continuous Improvement
- Just in time
- Impact of Poor Quality
- Cost of Quality
In this section we looked at the concepts of Quality, which is defined as conformance to your requirements. We analyzed the various approaches towards modern-day quality thinking, including Zero Defects and Kaizen. We also looked at how quality assurance should be a managerial initiative and how quality control is more of a technical function, involving the use of tools such as the Quality Control Chart.
In the next section, we will go into Project Human Resource Management
Ook! Time for a banana!