Professional Documents
Culture Documents
Om II Unit III Notes
Om II Unit III Notes
7 tools of quality
The ability to identify and resolve quality-related issues quickly and efficiently is essential to anyone
working in quality assurance or process improvement. But statistical quality control can quickly get complex
and unwieldy for the average person, making training and quality assurance more difficult to scale.
Thankfully, engineers have discovered that most quality control problems can be solved by following a few
key fundamentals. These fundamentals are called the seven basic tools of quality.
With these basic quality tools in your arsenal, you can easily manage the quality of your product or process,
no matter what industry you serve.
Learn about these quality management tools and find templates to start using them quickly.
Where did the quality tools originate?
Kaoru Ishikawa, a Japanese professor of engineering, originally developed the seven quality tools
(sometimes called the 7 QC tools) in the 1950s to help workers of various technical backgrounds implement
effective quality control measures.
At the time, training programs in statistical quality control were complex and intimidating to workers with
non-technical backgrounds. This made it difficult to standardize effective quality control across operations.
Companies found that simplifying the training to user-friendly fundamentals—or seven quality tools—
ensured better performance at scale
Today, these quality management tools are still considered the gold standard for troubleshooting a variety of
quality issues. They’re frequently implemented in conjunction with today’s most widely used process
improvement methodologies, including various phases of Six Sigma, TQM, continuous improvement
processes, and Lean management.
What are the 7 basic quality tools?
1. Stratification
2. Histogram
3. Check sheet (tally sheet)
4. Cause and effect diagram (fishbone or Ishikawa diagram)
5. Pareto chart (80-20 rule)
6. Scatter diagram
7. Control chart (Shewhart chart)
1. Stratification
Stratification analysis is a quality assurance tool used to sort data, objects, and people into separate and
distinct groups. Separating your data using stratification can help you determine its meaning, revealing
patterns that might not otherwise be visible when it’s been lumped together.
Whether you’re looking at equipment, products, shifts, materials, or even days of the week, stratification
analysis lets you make sense of your data before, during, and after its collection.
To get the most out of the stratification process, consider which information about your data’s sources may
affect the end results of your data analysis. Make sure to set up your data collection so that that information
is included.
2. Histogram
Quality professionals are often tasked with analyzing and interpreting the behavior of different groups of
data in an effort to manage quality. This is where quality control tools like the histogram come into play.
The histogram represents frequency distribution of data clearly and concisely amongst different groups of a
sample, allowing you to quickly and easily identify areas of improvement within your processes. With a
structure similar to a bar graph, each bar within a histogram represents a group, while the height of the bar
represents the frequency of data within that group.
Histograms are particularly helpful when breaking down the frequency of your data into categories such as
age, days of the week, physical measurements, or any other category that can be listed in chronological or
numerical order.
6. Scatter diagram
Out of the seven quality tools, the scatter diagram is most useful in depicting the relationship between two
variables, which is ideal for quality assurance professionals trying to identify cause and effect relationships.
With dependent values on the diagram’s Y-axis and independent values on the X-axis, each dot represents a
common intersection point. When joined, these dots can highlight the relationship between the two
variables. The stronger the correlation in your diagram, the stronger the relationship between variables.
Scatter diagrams can prove useful as a quality control tool when used to define relationships between quality
defects and possible causes such as environment, activity, personnel, and other variables. Once the
relationship between a particular defect and its cause has been established, you can implement focused
solutions with (hopefully) better outcomes.
LCLR = D3 R̅
Here the factors A2, D4 and D3 depend on the number of units per sample. Larger the number, the close the
limits. The value of the factors A2, D4 and D3 can be obtained from Statistical Quality Control tables.
However for ready reference these are given below in tabular form.
As long as X and it values for each sample are within the control limits, the process is said to be in statistical
control.
where d2 is a factor, whose value depends on number of units in a sample. Its value is seen from S.Q.C.
Tables 63.1.
Process Out of Control:
After computing the control limits, the next step is to determine whether the process is in statistical control
or not. If not, it means there is external causes that throws the process out of control. This cause must be
traced and removed so that the process may return to operate under stable statistical conditions.
The various reasons for the process being out of control may be:
(i) Faulty tools,
(ii) Sudden significant change in properties of new materials in a new consignment,
(iii) Breakdown of lubrication system,
(iv) Faults in timing of speed mechanisms etc.
Tracing of these causes is sometimes simple and straight forward but when the process is subject to the
combined effect of several external causes, then it may be lengthy and complicated business.
Process in Control:
If the process is found to be in statistical control, a comparison between the required specifications and the
process capability may be carried out to determine whether the two are compatible. Should the specified
tolerances prove to be too tight for the process capability?
There are three possible alternatives:
(a) Re-evaluate the specifications. Whether the tight tolerances are actually needed or they can be relaxed
without affecting quality.
(b) If relaxation in specifications is not allowed then a more accurate process is required to be selected.
(c) If both the above alternatives are not acceptable then 100% inspection is carried out to trace out the
defectives.
Example 1:
The table 63.2 give record of 5 measurements per sample from lot size of 50 for the critical dimension of
jeep valve stem diameter taken every hour, (i) Compare the control limits, make plot and explain plotting
procedure, (ii) Interpret plot, make decision regarding quality of product, process control and cost of
inspection.
From S.Q.C. table 63.1 the values of A2, D4 and D3 can be recorded from the 5 measurement sample column.
A2 = 0.58
and
D3 = 0
D4 = 2.11.
Now charts for X̅ and R are plotted as shown in Fig. 65.3 taking abscissa as sample number and ordinate as
X̅ and R. X̅ and R charts must be drawn one over the other as shown, i.e. R chart must be exactly under X̅
chart.
What the X̅ and R charts tell?
Sometimes X̅ chart does not give satisfactory results. This may occur due to old machine, or worn out parts
or misalignment or where processing is inherently quite variable. Here the “Range” chart is used as an
additional tool to control.
The purpose of this chart is to have constant check over the variability of the process. Process variability
demonstrated in the figure shows that though the mean or average of the process may be perfectly centred
about the specified dimension, excessive variability will result in poor quality products.
The use of R-chart is called for, if after using the X̅ charts, it is found that it frequently fails to indicate
trouble promptly.
The R-chart does not replace the X̅ -chart but simply supplements with additional information about the
production process.
The R-chart is also used for high precision process whose variability must be carefully held within
prescribed limits. Similarly many electro-chemical processes such as plating, and micro chemical biological
production, such as fermentation of yeast and penicillin require the use of R- chart because unusual
variability is quite inherent in such process.
Control Charts for Attributes:
The X̅ and R control charts are applicable for quality characteristics which are measured directly, i.e., for
variables. There are instances in industrial practice where direct measurements are not required or possible.
Under such circumstances, the inspection results are based on the classification of products as being
defective or not defective, acceptable as good or bad accordingly as that product confirms or fails to confirm
the specified specification.
In manufacturing, sometime it is required to control burns, cracks, voids, dents, scratches, missing and
wrong components, rust etc. Here, we inspect products only as good or bad but not how much good or how
much bad. Furthermore, there are many quality characteristics that come under the category of measurable
variables but direct measurement is not taken for reasons of economy.
These products are inspected with GO and NOT GO gauges. Again under this type also, our aim is to tell
that whether product confirms or does not confirm to the specified values. Quality characteristics expressed
in this way are known as attributes.
The various control charts for attributes are explained as under:
1. Attribute Charts for Defective Items: (P-Chart):
This is the control chart for percent defectives or for fraction defectives. This is used whenever the quality
characteristics are expressed as the number of units confirming or not confirming to the specified
specifications either by visual inspection or by ‘GO’ and ‘NOT GO’ gauges.
The Centre Line Value:
It is denoted by P̅ (P bar) and may be defined as the ratio between the total number of defective (non-
conforming) products observed in all the samples combined and the total number of products inspected. For
example, 15 products are found to be defective in a sample of 200, then 15/200 is the value of P̅.
Fraction and Percent Defectives:
The fraction defective value is represented in a decimal as proportion of defectives out of one product, while
percent defective is the fraction defective value expressed as percentage. As in the above example, fraction
defective of 15/200 = 0.075, and percent defective will be 0.075 x 100 = 7.5%.
Standard Deviation:
The standard deviation for fraction defective denoted by σ P is calculated by the formula.
Mostly the control limits are obtained on the basis of about 20-25 samples to pick up the problem and
standard deviation from the samples is calculated for further production control.
Example 3:
The table shows that successive lots of spindle are coming out of the machine. The spindles are subject to
inspection for burrs. The spindles are inspected in samples of 100 each.
Presence of a single or more burrs discriminates the value to be as defective. Compute and construct the
chart.
where,
X = process average
LSL = Lower Specification Limit
USL = Upper Specification Limit
σest = Process Standard Deviation
Step 7: Monitor The Process
The last step is to continually monitor the process and keep updating the SPC chart. Regular monitoring of a
process can provide proactive responses rather than a reactive response when it may be too late or costly.
Uses of SPC Charts
SPC charts are used for continuous improvement of a process using a number of techniques. There are
a number of ways SPC charts can help business analysts, but the most important ones are as follows:
Find and correct problems as soon as they occur
Predict the expected outcomes of a process
Determine whether a process is in a stable condition
Provide information on which areas to prioritize on to improve the process
steps to Creating an X-bar and R Control Chart
Once you decide to monitor a process and after you determine using an ¯XX¯ & R chart is appropriate, you
have to construct the charts.
This is not difficult and by following the 8 steps below you will have a robust way to monitor the stability of
your process.
1. Determine Sample Plan
Determine the sample size, n, and frequency of sampling. Consider the cost of sampling, required resources,
and balance with minimizing time (and produced units) between measurements. Of course, more samples
and more frequent measurements is better statistically.
2. Collect initial set of samples
Shewhart recommended 100 individual units in 25 samples of 4 each. Basically, we want enough samples to
create reasonable estimates of the mean values of ¯X and R. Keep the data in time sequence following the
time of the sample selection, which should be in the same order of manufacture.
3. Calculate ¯X
Calculate the average for each set of samples. This is the ¯X for each sample.
4. Calculate R
Calculate the range of each set of samples. This is the difference between the largest and smallest value in
the sample.
5. Calculate ¯¯X
Calculate the average of the ¯X’s. This is the centerline of the ¯X control chart.
6. Calculate R
Calculate the average of the R values. This is the centerline of the R control chart.
7. Calculate Control Limits
First, calculate the R chart limits.
Note: for a sample size, n, of less than 7 the lower limit is zero.
The D3 and D4 constants are found in tables or calculated with the following formulas
where,
and,
where,
8. Plot the data
With the control limits in place, gather samples, and plot the data. Look for special or assignable causes and
adjust the process as necessary to maintain a stable and in control process.
Pareto diagrams
Pareto chart is also called a Pareto diagram and Pareto analysis. It is named for the Italian
economist Vilfredo Pareto, who described Pareto principle, according to which roughly 80% of the
outcomes come from 20% of the conditions, for many events. This assumption is used in calculations of
business profit or population of any country. Hence, it is a part of probability and statistics.
Firstly, determine the classifications you will use to arrange the items in the bar graph.
Choose an appropriate measurement such as frequency, quantity, cost and time.
Decide the time period for the Pareto chart for which the bar graph will be drawn, For example, for a
week, a month or a year.
Assemble the new data and already existing data as per the category and period of time.
Do the total of the measurements for all the categories.
Choose the proper scale for the measurements you have assembled. The total you made in the above
step will be the maximum value.
Now for each category construct the bars along with labelling it. The longest bar should be placed at
the leftmost place and shortest at the rightmost.
For the analysis of the revenue growth of the organisation with respect to the time period.
To choose for any specific data and work on it, in a broad set of data available.
To explain to other people the set of data you have.
For the analysation of population growth in a city or country or all over the world every year.
To check the global problems and focus on resolving the major one.
To check the major complaints coming from the public and resolve them on priority.
1. A cause-effect diagram cannot identify a root cause; it presents graphically the many causes that might
contribute to the observed effect.
2. It is a visual representation of the factors that might contribute to an observed effect that is being
examined.
3. The interrelationships among the possible causal factors are clearly shown. One causal factor may
appear in several places in the diagram.
4. The interrelationships are generally qualitative and hypothetical.
5. It focuses the attention of all team members on the specific problem at hand in a structured, systematic
way.
When one has completed the diagram, one should be able to start at any endpoint and read the diagram as
follows (using Figure 29 as an example): “Snow causes the road to be slippery. The slippery road causes
loss of control of the car.” Alternatively, one can start with the phenomenon being explained and read it
backward like this: “Control of the car was lost because the road was slippery. The road was slippery
because it had snow.” In a properly constructed diagram, reading any of the branches in this way should
make good sense. We may not have evidence on which cause was actually the culprit, but the statement
should make good logical sense.
All possible sources of causation need to be considered. There are at least four classes of causes that may
apply to any problem:
C-E diagrams are generally easier to read and appear more visually pleasing if the text is placed at the end of
the line as in Figure 37. Other users have placed the text on the line like Figure 38. Text on the line tends to
be harder to use and read, especially as more levels of subsidiary causes are added.
Step 5 (Part B): Less Desirable Placement
Step 6 (Part A): Add Subsidiary Causes for Each Cause
Add subsidiary causes for each cause already entered. Each of these causes is placed at the end of a line
which is drawn (1) to connect with the line associated with the factor that it causes and (2) parallel with
either the main area line or the central spine. Figure 39 is an amplification of the portion of a C-E diagram
introduced in Step 5. Note how the governor and throttle have been added as possible causes of the wrong
speed of the engine. Throttle malfunction may result from either of two causes: Faulty calibration or
defective linkage.
Keeping the lines parallel makes reading easier and the visual effect more pleasing. Clearly, when one is
actually working on a C-E diagram in a team meeting, one cannot always keep the lines neat and tidy. In the
final documentation, however, it is found that using parallel lines makes for a more satisfactory diagram. A
diagram composed of lines with random orientation like the following example is harder to read and looks
less professional.
Step 6 (Part B): Less Desirable Placement
A root cause has three characteristics that will help explain when to stop. First, it causes the event the team
had sought after—either directly or through a sequence of intermediate causes and effects. Second, it is
directly controllable. That is, in principle, team members could intervene to change that cause. In the engine
example, we have been using in this section, speed cannot be controlled directly. Control of speed is
dependent on proper functioning of the throttle and governor, but proper control with the throttle is
dependent on correct calibration and proper functioning of the linkage. The calibration and the linkage can
be controlled. They are root causes.
Third, and finally, as the result of the other two characteristics, if the theory embodied in a particular entry
on the diagram is proved to be true, then the elimination of that potential root cause will result in the
elimination or reduction of the problem effect that we were trying to explain.
Step 8: Check Logical Validity of Each Causal Chain
Check the logical validity of each causal chain. Once the entire C-E diagram is complete, it is wise to start
with each potential root cause and “read” the diagram forward to the effect being explained. Be sure that
each causal chain makes logical and operational sense. Consider the following example, which is a portion
of a C-E diagram seeking to explain errors in an order-entry process. One main area of errors concerns
errors in the part numbers. Sales representatives look up the part in a catalog and enter the part number on an
order form. The information from the form is then keyed into a database.
Start with the proposed root cause “keying error.” Then read it as follows: “Keying errors cause fatigue
which causes the wrong part numbers…” Once we try to read the diagram, the problem becomes clear.
Keying errors do not cause fatigue; fatigue causes keying errors, and the diagram should be reorganized as
follows.
This redrawn diagram places fatigue, format, and training as root causes of three different intermediate
causes of the wrong part numbers —misreading the catalog, entering the data on the form improperly, and
keying the data improperly. Because these now trace out logical causal chains, it is easier to devise effective
ways of testing the theories. For example, form formats which cause problems in keying may differ from
those which create problems in the original pencil entry.
The general “lack of training” cause on the original diagram is normally a good danger sign that the causal
chain needs to be checked. Lack of training in reading the catalog will create reading errors, but if the errors
come at the keying stage, no amount of training on use of the catalog will do any good. Whenever one sees
“lack of training” (or lack of anything else for that matter) on a C-E diagram, one should ask two questions.
First, exactly which skill is training lacking in? And second, how does that lack cause the factor being
explained at the moment? As we saw in our example here, answers to those questions may help identify
missing intermediate causal factor and causal relationships that are stated backward.
Step 9: Check for Completeness
As discussed more fully in the interpretation section, check for the following:
Main branches with fewer than three causes
Main branches with substantially fewer causes than most others
Main branches that go into less detail, with fewer levels of subsidiary causes than do the others
Main branches that have substantially more causes than most of the others
The existence of one of these conditions does not automatically mean a defect in the diagram; it merely
suggests that further investigation is warranted. At this point, it is also good to double check that the 4 W’s,
5 M’s, and/or 5 P’s are considered as appropriate.
When to Use Cause and Effect Diagrams
Formulating Theories
The chief application of the cause-effect diagram is for the orderly arrangement of theories about the causes
of the observed quality problem that the team is assigned to resolve. Once the theories are well understood
and ordered, then the team will use its best collective judgment to identify those theories which should be
tested. The final objective of the characterize step is the identification of the primary root cause or causes of
the team’s problem.
There are also other opportunities for organizing theories. The team may want to know why some part of the
process works better than other parts. For example:
Why does automobile A obtain ten percent better mileage per gallon than all other similar vehicles
tested?
Why is the productivity of assembly line B always higher than the productivity of the other lines?
Designing for Culture
During the Improve step, the cause-effect diagram may also be useful for the team in considering the
cultural impact of its proposed remedy. A cause-effect diagram can sometimes be helpful in thinking
systematically about the resistance that the proposed solution is likely to meet. If the phenomenon to be
explained is resistance to the proposed remedy, then the team can construct a cause-effect diagram to help
identify the most important resistances it will need to address.
How to Interpret Cause and Effect Diagrams
The Result
The cause-effect diagram does not provide an answer to a question, as some other tools do. Its main value is
to serve as a vehicle for producing, in a very focused manner, a list of all known or suspected causes which
potentially contribute to the observed effect. At the time of generating the cause-effect diagram, it is not
usually known whether these causes are responsible for the effect or not.
A well-prepared cause-effect diagram is a superb vehicle for helping to reach a common understanding of a
complex problem, with all its elements and relationships clearly visible at whatever level of detail is
required. The output of the diagram can be leveraged by prioritizing potential causes or theories for further
investigation.
The Difference between Theory and Fact
We have noted that cause-effect diagrams present and organize theories. Only when theories are tested with
data can we prove causes of observed phenomena. The cause-effect diagram helps organize the search for
the causes, but it does not identify the causes. Other tools, such as Pareto analysis, scatter diagrams, and
histograms, will be used to analyze data to establish the causality empirically.
Checking for Completeness
As a cause-effect diagram is constructed, it should be reviewed for completeness. There can be no exact
rules for this review, but some guidelines are helpful. Some of these are discussed in more detail in the
construction section.)
Be certain that you have at least asked how each of the 4 W’s and each of the 5 M’s or 5 P’s might apply to
the effect.
Generally, each main branch of the diagram will have at least three or four additional branches. If one does
not, further consideration of that branch may be advisable to verify that it has been understood fully.
If some main branches have substantially fewer causes attached to them, or if the causes on them do not go
back as many steps in the causal chain, you may not have as full an understanding of that element of the
process as you do of the others. It may be appropriate to seek theories from additional persons familiar with
that element of the process.
If some branches appear overloaded with causes compared to the others, consider whether they might be
most appropriately divided into two or more main branches.
Verify that the cause at the end of each causal chain is potentially a root cause. Such a potential root cause
will normally satisfy three conditions. (1) You can trace a logical causal relationship from that cause,
through all its intermediate causes, to the final effect being explained. (2) That cause is, in principle, directly
controllable. (3) Therefore, if shown to be true, that cause could be eliminated, and the effect would
disappear or be reduced.
Potential Pitfalls and Problems in Interpretation
The most serious possible misinterpretation of a cause-effect diagram is to confuse this orderly arrangement
of theories with real data. The C-E diagram is a powerful and useful way to develop theories, display them,
and test their logical consistency. It is no substitute for empirical testing of the theories.
We will discuss in more detail later the need to test each causal relation in the C-E diagram for logical
consistency. Failure to make those checks can greatly reduce the usefulness of the diagram and often lead to
the waste of valuable time collecting and analyzing the wrong information.
Another common pitfall is to begin construction of the diagram before the symptoms have been analyzed as
thoroughly as existing information will permit. In such cases, the effect being explained may be so general
and ill-defined that the team will have a hard time focusing and the resulting diagram may be unnecessarily
large, complex, and difficult to use. A clear and precisely articulated effect will produce more relevant
theories, better causal relationships, and a more effective model for the selection and testing of theories.
A final pitfall is to limit the theories that are proposed and considered. While the symptom being explained
should be as precisely defined as possible, the team must seek to develop just as many theories as possible
about its causes. If a team does not develop a wide-ranging set of theories, they might miss their most
serious root cause.
What Is Done Next?
With a complete and logical set of theories in hand, the team will now want to discover which are the
principal root causes. This structured approach to identify theories allows investigation of those of
importance rather than wasting time on trivial theories. One or more of these theories will be selected for
testing, collect the data needed for the test, and apply one or more other tools to these data to either confirm
or deny the tested theories.
Scatter diagrams
A scatter diagram is one of seven core tools in project management. It is used to plan and monitor operations
to improve quality-related issues in an organization. Scatter diagrams are graphical statistical tools. They are
simple to use and help in improving business processes.
While typical charts and graphs use lines or bars to represent data, scatter diagrams use dots. At first glance,
this may be confusing, but scatter diagrams are easy to understand if you take the time.
What is a scatter diagram?
A scatter diagram is a graph that shows the association between two variables for a collection of numerical
data. It depicts the link between a process component on one axis and the quality fault on the other to help
with process optimization.
A scatter diagram demonstrates the relationship between a change in a dependent variable Y in response to a
change in a corresponding independent variable X. How do we know which is the response and independent
variable?
In general, the independent variable tries to explain or anticipate an observed outcome. The response
variable measures the outcome. When creating the graph, the points will fall along a line or curve if the
variables are correlated.
A scatter chart might be helpful when one variable is measurable, but the other is not. You can forecast the
behavior of the dependent variable based on the independent variable after establishing how the variables
are connected.
When would you use a scatter plot?
A scatter plot is an excellent tool for planning and measuring quality when:
Remove barriers, find clarity, exceed goals
Anything is possible with the most powerful work management software at your fingertips
Try for free
1. You have numerical data that is paired
2. Each value of your independent variable might have several values for your dependent variable
3. Objectively deciding whether a cause and effect are linked
4. Assessing if two seemingly similar outcomes are caused by the same thing
5. One variable can be measured, but the other can't
6. Investigating hypotheses concerning cause-and-effect interactions
7. Looking for the root of an issue that has been recognized
For instance, we may look at the pattern of plant height over time. You build the graph after selecting the
two variables: the plant's height and age. Once you've finished drawing the scatter diagram, you may
observe that as a plant grows older, it grows taller. This demonstrates a link between plant height and age.
The three types of scatter diagrams
The link between variables in scatter diagrams is indicated by the direction of the correlation on the graph. A
correlation in a scatter diagram occurs when two variables are determined to have a connection.
You can use a regression line to predict how a change in one variable will likely affect the value of a
dependent variable. You can find a correlation when a cause-and-effect link exists between both variables.
Three types of correlations in scatter diagrams are:
Positive correlation
If variables have a positive correlation, this signifies that when the independent variable's value rises, the
dependent variable's value rises as well. Consider this scatter diagram example:
As the weight of human adults increases, the risk of diabetes also increases. The pattern of observation in
this example would slant from the chart's bottom left to the upper right.
Negative correlation
In the negative correlation, when the value of one variable grows, the value of the other variable falls. The
dependent variable's value drops as the independent variable's value rises.
Here’s an example: When summer temperatures rise, sales of winter clothing declines. The pattern of
observation in this example would slant from the top left to the bottom right of the graph.
No correlation
The "no correlation" type is used when there's no potential link between the variables. It's also known as
zero correlation. The two variables plotted aren't connected in any way.
The area of land and air quality index, for example, have no relationship. As an area grows, there is no effect
on the air quality. These two variables have no association, and the observations will be dispersed all over
the graph.
A fundamental observation to remember when studying correlation is that a link between two variables does
not guarantee causation. Stay mindful that correlation does not always indicate causation. A correlated
relationship may happen for any of the following reasons:
The causal relationship being reversed
A third variable causing it
Accidents and coincidences
The advantages and disadvantages of scatter diagrams
Advantages of scatter diagrams include:
1. Patterns are easy to spot in scatter diagrams
2. A scatter diagram is easy to plot with two variables
3. Scatter diagrams are an effective way to demonstrate non-linear patterns
4. Scatter diagrams make it possible to determine data flow range, such as the maximum and minimum
values
5. Plotting scatter diagrams helps with better project decisions
6. Scatter diagrams help uncover the underlying root causes of issues
7. They can objectively assess if a given cause and effect are connected
Disadvantages of scatter diagrams include:
1. Reading scatter diagrams incorrectly may lead to false conclusions that one variable caused the other,
when both may have been influenced by a third
2. A relationship in a scatter diagram may not be apparent because the data does not cover a wide
enough range
3. Associations between more than two variables are not shown in scatter plots
4. Scatter diagrams cannot provide the precise extent of association
5. A scatter plot does not indicate the quantitative measure of the relationship between the two variables
Scatter diagram example
A scatter diagram can be applied to any data model with two variables and their respective numerical data.
Let's look at a scatter diagram example.
We'll examine the number of workplace accidents happening at a factory. The two variables are the number
of shift hours and the number of accidents. We will produce our scatter diagram based on the following data
provided:
The independent or control variable on the horizontal axis are shift hours, while the dependent variable on
the vertical axis is the number of accidents.
After drawing the scatter diagram, we see that the number of accidents increases as the number of shift
hours increases. This illustrates a positive correlation between the two.
Scatter plots do not always have a controlling parameter. It is possible to have two independent variables. In
such a situation, any axis can represent either independent variable.
Scatter diagrams in PMP
As part of the Project Management Professional (PMP) certification test preparation, it is crucial to
understand scatter diagrams in PMP terms. To be certified, aspiring project managers need to learn to create
and gain insights from scatter plot analysis.
The potential PMP candidate may be asked to evaluate data using scatter diagrams or choose the best quality
control tool to employ in a given circumstance. Although there are no specific scatter diagram examples for
the PMP exam, a general understanding of the tool is expected.
How to create a scatter diagram
Creating a scatter diagram can be broken into the following five steps:
1. Identify variables: Identifying the independent and dependent variables is the first step in creating a
scatter diagram. Find out which is the control variable affecting the dependent variable. Use variables
that are quantitative and objective. For example, say you are observing the time taken by a vehicle to
reach its destination at different speeds. The speed of the vehicle is the independent or control
variable, while the distance is the dependent variable. Remember, it is also possible for both variables
to be independent.
2. Pull data: After you have determined both variables, gather data from the variables either by
witnessing the process or using digital sources and tools, such as analytics, maintenance systems,
automation, or mobile audit software.
3. Build the scatter plot: Once you've gathered your data, use a spreadsheet or scatter plot tool to
develop your scatter plot by connecting the dots representing each collection of numerical data. In our
earlier example, we would see how many hours it takes to get to the destination at different speeds.
The speeds are plotted on the x-axis, while the hours are plotted on the y-axis.
4. Determine the type of correlation: After plotting the dots on the scatter diagram, analyze and
determine the correlation between the two variables. The data trend could be upward, downward, or
undefined. The correlation between the two variables may be positive or negative, or there may be no
correlation. In our example, the correlation is negative as the number of hours declines with increased
speed.
5. Conduct a scatter plot analysis: We may come to several conclusions depending on how the scatter
plot turns out. Confirm your conclusions by using scatter plots in conjunction with other root cause
analysis methods. The key to fixing problems and building lasting remedies is to look at them from
various perspectives.
Getting started with scatter diagrams
Scatter diagrams help discover correlations between variables and guide quality control in project
management. They're a crucial part of the PMP certification exams and help project managers make better
decisions.
Are you looking to improve your business processes and supercharge your project management? Use scatter
diagrams to compare elements and confirm your conclusions. Get a two-week free trial of Wrike’s project
management software to keep your ongoing projects organized.
Run Chart
What is a run chart?
A run chart is used to study collected data for trends or patterns over a specific period of time. A run chart
will help you:
Monitor data over time to detect trends, shifts, or cycles
Compare a measure before and after the implementation of solution to measure impact
Focus attention on vital changes, not normal variation
Track useful information for predicting trends
The run chart is a running record of a process over time:
The vertical axis represents the process being measured
The horizontal axis represents the units of time by which the measurements are made
The centerline of the chart is the mean or average
A run is defined as one or more consecutive data points on the same side of the mean line.
How to create a run chart
1. Choose which data you will measure and track
2. Gather data: Generally, collect 20-25 data points, with which you can detect meaningful patterns over time
3. Create a graph on which you can plot your data (y axis, or vertical line) over time (x axis, or horizontal line)
4. Plot the data
5. Interpret the chart: Focus on the vital changes or meaningful trends/patterns, rather than each and every data
variation; keep reading for interpretation tips
Using a run chart to test for special causes
Histograms
Frequency distribution shows how often each different value in a set of data occurs. A histogram is the most
commonly used graph to show frequency distributions. It looks very much like a bar chart, but there are
important differences between them. This helpful data collection and analysis tool is considered one of
the seven basic quality tools.
WHEN TO USE A HISTOGRAM
Use a histogram when:
The data are numerical
You want to see the shape of the data’s distribution, especially when determining whether the output of a
process is distributed approximately normally
Analyzing whether a process can meet the customer’s requirements
Analyzing what the output from a supplier’s process looks like
Seeing whether a process change has occurred from one time period to another
Determining whether the outputs of two or more processes are different
You wish to communicate the distribution of data quickly and easily to others
Histogram Example
HOW TO CREATE A HISTOGRAM
1. Collect at least 50 consecutive data points from a process.
2. Use a histogram worksheet to set up the histogram. It will help you determine the number of bars, the
range of numbers that go into each bar, and the labels for the bar edges. After calculating W in Step 2 of
the worksheet, use your judgment to adjust it to a convenient number. For example, you might decide to
round 0.9 to an even 1.0. The value for W must not have more decimal places than the numbers you will
be graphing.
3. Draw x- and y-axes on graph paper. Mark and label the y-axis for counting data values. Mark and label
the x-axis with the L values from the worksheet. The spaces between these numbers will be the bars of the
histogram. Do not allow for spaces between bars.
4. For each data point, mark off one count above the appropriate bar with an X or by shading that portion of
the bar.
HISTOGRAM ANALYSIS
Before drawing any conclusions from your histogram, be sure that the process was operating normally
during the time period being studied. If any unusual events affected the process during the time period of
the histogram, your analysis of the histogram shape likely cannot be generalized to all time periods.
Analyze the meaning of your histogram's shape. Typical histogram shapes and what they mean are covered
below.
HISTOGRAM WORKSHEET EXAMPLE
TYPICAL HISTOGRAM SHAPES AND WHAT THEY MEAN
Normal Distribution
A common pattern is the bell-shaped curve known as the "normal distribution." In a normal or "typical"
distribution, points are as likely to occur on one side of the average as on the other. Note that other
distributions look similar to the normal distribution. Statistical calculations must be used to prove a normal
distribution.
It's important to note that "normal" refers to the typical distribution for a particular process. For example,
many processes have a natural limit on one side and will produce skewed distributions. This is normal—
meaning typical—for those processes, even if the distribution isn’t considered "normal."
Skewed Distribution
The skewed distribution is asymmetrical because a natural limit prevents outcomes on one side. The
distribution’s peak is off center toward the limit and a tail stretches away from it. For example, a distribution
of analyses of a very pure product would be skewed, because the product cannot be more than 100 percent
pure. Other examples of natural limits are holes that cannot be smaller than the diameter of the drill bit or
call-handling times that cannot be less than zero. These distributions are called right- or left-skewed
according to the direction of the tail.
Double-Peaked or Bimodal
The bimodal distribution looks like the back of a two-humped camel. The outcomes of two processes with
different distributions are combined in one set of data. For example, a distribution of production data from a
two-shift operation might be bimodal, if each shift produces a different distribution of results. Stratification
often reveals this problem.
Comb Distribution
In a comb distribution, the bars are alternately tall and short. This distribution often results from rounded-off
data and/or an incorrectly constructed histogram. For example, temperature data rounded off to the nearest
0.2 degree would show a comb shape if the bar width for the histogram were 0.1 degree.
Truncated or Heart-Cut Distribution
The truncated distribution looks like a normal distribution with the tails cut off. The supplier might be
producing a normal distribution of material and then relying on inspection to separate what is within
specification limits from what is out of spec. The resulting shipments to the customer from inside the
specifications are the heart cut.
Flow charts
WHAT IS A FLOWCHART?
Also called: process flowchart, process flow diagram
Variations: macro flowchart, top-down flowchart, detailed flowchart (also called process map, micro map,
service map, or symbolic flowchart), deployment flowchart (also called down-across or cross-functional
flowchart), several-leveled flowchart
A flowchart is a picture of the separate steps of a process in sequential order. It is a generic tool that can be
adapted for a wide variety of purposes, and can be used to describe various processes, such as a
manufacturing process, an administrative or service process, or a project plan. It's a common process
analysis tool and one of the seven basic quality tools.
Elements that may be included in a flowchart are a sequence of actions, materials or services entering or
leaving the process (inputs and outputs), decisions that must be made, people who become involved, time
involved at each step, and/or process measurements.
WHEN TO USE A FLOWCHART
To develop understanding of how a process is done
To study a process for improvement
To communicate to others how a process is done
When better communication is needed between people involved with the same process
To document a process
When planning a project
FLOWCHART BASIC PROCEDURE
Materials needed: Sticky notes or cards, a large piece of flipchart paper or newsprint, and marking pens.
1. Define the process to be diagrammed. Write its title at the top of the work surface.
2. Discuss and decide on the boundaries of your process: Where or when does the process start? Where or
when does it end? Discuss and decide on the level of detail to be included in the diagram.
3. Brainstorm the activities that take place. Write each on a card or sticky note.
4. Arrange the activities in proper sequence.
5. When all activities are included and everyone agrees that the sequence is correct, draw arrows to show the
flow of the process.
6. Review the flowchart with others involved in the process (workers, supervisors, suppliers, customers) to
see if they agree that the process is drawn accurately.
FLOWCHART CONSIDERATIONS
Don’t worry about drawing the flowchart the "right way." Ultimately, the right way is the way that helps
those involved understand the process.
Identify and involve in the flowcharting process all key people involved with the process. This includes
suppliers, customers, and supervisors. Involve them in the actual flowcharting sessions by interviewing
them before the sessions and/or by showing them the developing flowchart between work sessions and
obtaining their feedback.
Do not assign a "technical expert" to draw the flowchart. People who actually perform the process should do
it.
FLOWCHART EXAMPLES
1. High-Level Flowchart for an Order-Filling Process
2. Detailed Flowchart
Decision based on a question. The question is written in the diamond. More than one arrow goes
out of the diamond, each one showing the direction the process takes for a given answer to the question.
(Often the answers are "yes" and "no.")
Delay or wait
Link to another page or another flowchart. The same symbol on the other page indicates that the
flow continues there.
Input or output
Document