Review 1 Flashcards
Name the 3 categories of the Kano model
Dissatisfiers, Satisfiers, and Delighters
Name 3 strategies for ensuring quality in a process
- Quality by Design
- Quality by Process Monitoring and Control
- Quality by Self-Check and Verification
Name 3 Levels of Error-Proofing
Prevention, Facilitation (mitigate errors), and Detection
Which acronym will help you remember the 8 types of waste?
TIM WOODS
What is the purpose of a Failure Modes and Effects Analysis (FMEA)?
To try to anticipate and eliminate ways that a process could fail.
When is a Kaizen exercise most appropriate?
When the primary drivers of a process’s performance are already known.
When is a DMAIC project most appropriate?
When an improvement effort is needed and the primary drivers of performance are not yet known
What type of project is needed when a process is so broken that it should be redesigned from the ground up?
DMADV
What is TAKT Rate?
the rate of customer demand
What is Throughput Rate?
the rate at which the process can produce output
Name 4 types of the costs of Quality
- Appraisal Costs
- Prevention Costs
- Internal Failure Costs (scrap, rework, repair)
- External Failure Costs (loss of customer good will)
What is Rolled Throughput Yield (RTY)?
the average percentage of yield with no defects
What are the 3 key elements of quality?
- Customer
- Process
- Employee
What is the function of a VOC to CTQ matrix?
it helps to translate “critical customer requirements” into CTQs
What is a z-score?
the number of standard deviations which fit between the specification limits and the mean
Name the 3 key elements of Lean Management
- Purpose
- Process
- People
Lean Focuses on which 3 factors which inhibit performance?
- Waste
- Variability
- Inflexibility
Name the 5 steps of the 5-S system to reduce waste and optimize productivity via an orderly workplace
Sort, Set in Order, Shine, Standardize, Sustain
Name 4 strategies to eliminate non-value-add steps in a process
Simplify,
Eliminate,
Automate,
Standardize
What are the most common causes of defects?
- Inappropriate operating procedures
- Excessive variation in operations
- Defective raw material
- Human or Machine error
What is Kanban?
a scheduling system for Lean and just-in-time production
Process Cycle Efficiency (or Value Added Ratio) can be described by which formula?
(value add time) / (Total cycle time)
What is “overall equipment effectiveness (OEE) and which 3 factors is it based on?
OEE is a hierarchy of metrics that provide a KPI as an indicator of success. It’s based on Availability, Performance, and Quality.
What is the purpose of the Theory of Constraints?
to indicate how to use available resources most effectively
What is “The 5-Why Analysis”?
it is a simple root cause analysis tool. Ask “why?” 5 times to each root cause.
What is Process Stability?
it defines how consistent are the outputs of a process
What is Process Capability?
The ability of a process to meet specifications
Which should be established first? Stability or Capability?
Stability should be established first
Name 3 types of Location Variations
Bias, Stability, and Linearity
Name 2 types of Spread Variations
Repeatability and Reproducability
Name two useful tools for Measurement System Analysis
- Gage Run Chart
2. Gage R&R Study
In MSA, what does the number of trials and parts generally reflect?
It reflects the criticality of the measurements. More critical equals more trials and parts.
How should MSA be set up?
- The measurement tools should be the same as those used in the production process.
- Number the parts in advance.
- Measure the parts in a random order.
In MSA, what is the minimum number of trials and appraisers?
At least two trials are required and two appraisers so that both Reproducibility and Repeatibility can be measured.
Name a statistical tool that is useful in MSA for Attributes
Attribute Agreement Analysis
What does it mean if the Potential Process Capability (Cp) is >= 1 ?
It means that the process is “potentially capable”.
If the Actual Process Capability is Cpk, then what is the Sigma Level of the process?
Sigma Level = 3 X Cpk
What are the 5 Key Steps of the Analyze Phase?
- List the x’s that may impact Y
- Organize the potential x’s
- Select the most likely Key x’s
- Develop a Data Collection Plan
- Prove which are the key x’s in Y = f(x)
In order to help identify the key x’s in a process, name 3 graphs for when Y is continuous and X is discrete.
- a histogram
- a box plot
- a dot plot
In order to help identify the key x’s in a process, name a good Normality Test for when Y is continuous and X is discrete.
the Anderson-Darling test
In order to help identify the key x’s in a process, name 3 Compare-Means tests for when Y is continuous and X is discrete.
- 1-Sample t test
- 2-Sample t test
- ANOVA
Name two Variance tests
- Chi-Square
2. the Bonett method (if data is not normal)
Name two tests for equal variance
- Bartlett’s Test
2. Levene’s test (if data is not normal)
If data is not normal, you will typically compare medians instead of means. Name 3 tests used to compare medians.
- Wilcoxon
- Mann-Whitney
- Kruskal-Wallis
What if both X and Y are continuous? How will you compare values?
using a Scatter plot or a Fitted Line plot
What if both X and Y are DISCRETE? How will you compare values?
using a Pareto Chart
How would you test for correlation or independence?
Chi-square contingency tables
What is a type-1 error?
Rejecting the NULL hypothesis when it is in fact true
What is a type-2 error?
Failure to reject the NULL hypothesis when it it false.
What is Beta in error probability?
The probability of a type-2 error happening
define the “Power” of a test
Power = 1 - Beta It is the probability of correctly rejecting the null hypothesis when it is false. You want the Power of a test to be as close to 1 as possible.
How is the Anderson-Darling Test used?
It is a test for data normality. The null hypothesis is “the Data is Normally Distributed”. The alternate is “the data is not normal”. If the P-value is greater than 0.05, you fail to reject the null and you conclude that the data is normally distributed.
If you perform a test and discover that your data is not normally distributed, what are your options?
Compare medians using non-parametric tests such as:
- Wilcoxon
- Mann-Whitney
- Kruskal-Wallis
What would you use to compare the variances of non-normal data?
Levene’s Test
What are the two key assumptions about the data when you are using ANOVA?
- The variances are equal (you must confirm this)
2. The data is normally distributed (confirm this too)
When testing for correlation, what is the null hypothesis?
That there is NO significant correlation
What 4 steps will you follow in the course of the Improve phase?
- Generate Potential solutions to the process problems
- Evaluate solution options
- Select the most impactful solutions
- Pilot and implement your changes
Name some tools that you would typically use in the course of the Improve phase
- Process Maps
- FMEA
- Mistake Proofing
- Pilot Testing
When is it appropriate to use “Response Surface Methods”?
When it is discovered that your regression function has to account for curvature. A second order or Quadratic model is then required.
What are the general goals of the Control Phase?
- Develop a control plan to monitor, control, and regulate performance
- Work with process owners to update procedures and develop/implement training plans
- Help Implement Changes
- Determine improved process capability and Sigma Level
- Validate the financial impact of the process improvements
- Celebrate once the champion signs off
What is a P-chart used for?
plots the proportion of defective units
What is an NP-chart used for?
plots the number of defective units
What is a C-Chart used for?
plots the number of defects
What is a U-chart used for?
plots the number of defects per unit
What is the Round Robin technique?
a brainstorming strategy where everyone gets a chance to put forth his/her ideas
What is the “anti-solution” technique?
a brainstorming strategy that involves ideas on how to make the problem worse rather than solve it. The ideas are then reversed to get the solution to the problem.
What is the “6-Hats” technique?
Each team wears one colored hat out of 6 and each color signifies the role he will play in the brainstorming session.
What is the “Nominal Group Technique”?
a structured form of brainstorming that results in the generation and prioritization of solutions
What is a “Noise Variable”?
an unknown or unaccounted for input variable that is the real reason your Y changed. – Solution: randomization
What is the solution when you have too little data in too short a time period?
Repetitions within each run
Name and describe the most powerful of the DOE techniques
Screen Designs – e.g. Plackett-Burman DOE, Taguchi. These typically have fewer runs but more factors (6-11). Very powerful for uncovering the main factors in a process.
Describe DOE Characterization Designs
3-5 factors – Fractional and Full Factorial DOEs. They typically have more runs but fewer factors and tend to provide more detail.
Describe DOE Optimization Designs
<= 3 factors –> Response Surface Method DOEs. Typically have both fewer runs and fewer factors. The Central Composite Design will be the most efficient RSM. These DOEs tend to provide a whole lot of detail.
Why are DOEs so powerful?
They provide a way to change many inputs at once to get faster yet reliable results regarding which inputs primarily change Y.
What do the R-squared and Adj R-squared tell you?
Whether or not you have too many terms in your equation for the amount of data that you collected.
What do you do if R-squared and Adj R-squared are more than roughly 4% apart?
Eliminate the term with the highest P-value from your equation
What are residuals?
Our error terms. They will tell us whether our equation works well for all predicted values or just for some of them.
What are two good Residual Analyses?
- Histogram
2. Residual vs Fits
When analyzing residuals on a histogram, what do we typically want to see?
a normal distribution. But on other plots, we want to see randomness.
What is the ultimate Goal of DOE?
To get the most information about the important factors of a process by using the fewest amount of runs
Which 4 items must be evaluated as part of the DOE analysis?
- P-values
- R-Square
- +/- 2 * S
- Residuals
In DOE analysis, what is the P-value used for?
it will reveal the significance of various terms in your equation
In DOE analysis, what is the R-square used for?
it describes the relationships of inputs to outputs
In DOE analysis, why calculate +/- 2 * S ?
it is an indication of the predictability of an equation
In DOE analysis, why evaluate residuals?
they are examples of violations of your analysis assumptions
What are the general goals of LEAN methodology?
The elimination of waste in order to improve efficiency, flow, and speed.
Define Value
Value is what customers need, expect, and are willing to pay for.
Define Value Stream
Processes, activities, and resources, including information, used to transform inputs into outputs that are saleable to customers
What is a bottleneck or constraint?
The slowest processing step of a value stream
What is Just-in-Time Flow?
Items or transactions should be produced, processed, or delivered at the pace of customer demand.
What are CTQs?
They are performance characteristics of a process or product or service that are critically important to customers
Why calculate the Cost of Poor Quality (CoPQ)?
Those costs are what you can expect to recoup and gain from a successful Lean Six Sigma project.
What are the most important goals of Value Stream Mapping?
To get a sense of 1) where value is added, and 2) where waste occurs.
What is Lead Time?
Lead Time = Throughput Time. Total time for an item or transaction to move through an entire process.
Define Cycle Time
The average time for a unit to be produced or processed.
An Attribute Agreement Analysis will try to measure which three things?
- Percent agreement between appraisers
- Percent agreement within appraisers
- Percent agreement with the expert
All agreement should be at least 90% if not higher.
Name two different types of descriptive statistics
- Measures of Central Tendency
- - Measures of Spread or Variation
What are the Four Ws of Stratification in the Measure Phase?
Who, What, Where, and When
What is the rule of thumb for when you want to use a histogram?
You need to have at least 45 data points.
What factor is most critical to ensure that your experiment has sufficient “Power”?
Sample Size
When dealing with a number of possible solutions to implement, which tool can help focus on the best solutions?
a Criteria Selection Matrix
Which chart is used to track the stability of the average value of a metric of interest?
Xbar-R Chart
Which graph demonstrates the conditions which would be sufficient to enable OCAP for the process?
Xbar Chart
Control Charts were developed by Shewhart to track data over time. To detect special cause variation, the control charts use what?
Center line and control limits
What types of process variation are the focus of Statistical Process Control?
Common cause and Special Cause variation
Special Cause variation falls into which two categories?
Assignable and Pattern
What is the technique/tool used to determine if Special Cause errors are occurring within the subgroups of SPC charts?
Range Charts
If the production is for higher volume and monitoring and the mean and variability is to be monitored for four machines and the characteristic to be monitored is Variable Data, which SPC chart is best to be selected?
Xbar-R Chart
T or F: a fundamental rule is that both Standard Deviation and Variance can be added.
False
In addition to a Control Plan, what else is developed at the completion of a LSS project so that those involved know what to do when critical process metrics move out of spec?
a Response Plan
The Control Limits width varies if the sample size varies for which type of chart?
P charts
Name 3 types of Planned Experiments
- Fractional Factorial
- Factorial
- RSM
Which Experimental Design typically has the fewest factors or input variables in it?
RSM
T or F: The Lean toolbox includes Design of Experiments
False
The method used by Kanban is to require a _____ before anything moves.
Signal
Why use a fractional factorial design instead of full factorial?
The reduction in the number of runs will reduce both the time and the cost associated with the experiments.
a Full Factorial experiment using a 3 level 3 factor approach has been proposed to test the viability of an extrusion machine experiment. How many treatment combinations will this approach involve?
27
When doing hypothesis testing on non-normal data, belts will use what to compare more than two sample proportions to each other?
a Mean-To-Mode analysis
It would be more likely than not for a belt conducting a regression analysis to find that the r2 value is smaller than what?
the absolute value of r
Which 3 phenomena might contribute to similar distributions having unequal variances?
- extreme tails
- outliers
- multiple modes
The relationship between a response variable and one or more independent variables is investigated and modeled using what?
ANOVA
How does one assess process proportion as opposed to evaluating a process with respect to a set target?
Process Proportion equals some desired value
Which technique is best used to compare a machine 1 average quality characteristic to the same quality characteristic of machine 2?
a 2-sample t-test
What has occurred when 2 inputs have a greater impact on a change in output than either of the inputs has by itself?
Interaction
To be an effective Lean Six Sigma practitioner, one must understand the difference between …..
Practical and Statistical Significance
Why might a belt see multiple modes in a graphical analysis while analyzing data?
The process may have experienced a significant change from one shift to another; there could have been a sizable measurement system error; There might have been a catastrophic failure of some sort.
Time is always the metric on the horizontal scale of what kind of chart?
A multi-var chart
Which 3 types of variation can impact a process?
Within, Between, and Temporal
What could be used to screen variables to analyze their relative impact on the output of concern?
a Multi-Vari chart
Some of the sources for different types of error that can be quantified using statistical analysis are:
- error in sampling
- bias in sampling
- error in measurement
In a normality test such as Anderson-Darling, what P-values would indicate that the data is normally distributed?
P >= 0.05
With Measurement System Analysis, we are concerned with two issues that impact the potential variability of the data. What are they?
Precision and Accuracy
What is defined as the difference between the observed and the expected values for a given set of data?
Measurement Bias
Appropriate measures means that measurements are what 4 things?
- Representative
- Sufficient
- Contextual
- Relevant
What kind of problem in the measurement system suggests that there is a lack of consistency in the measurement over time?
a stability problem
In a good measurement system, the most variation will be with part-to-part measurements. What should you do if the majority of variation is associated with the Gauge R&R, assuming the gauge is technically capable?
Focus on fixing the repeatability and reproducibility of the measurement device
What aspects of Measurement Systems Analysis (MSA) studies are applicable when the process that is used to measure does not damage the part?
Nondestructive variable gage R&R and Nested Study
When we compare short-term and long-term Capability, what is true of Cp?
Cp is better for the short term
Relative to a Design of Experiments, the term Colinear refers to variables being what??
a linear combination of each other
Which experimental design is most associated with the fewest number of factors in the design?
Response Surface design
Which techniques are typically considered to be part of the Lean toolbox?
- Poke-Yoke
- Kaizen
- 5s the work area
If a process has subgroups for variable data and the process runs for a long period of time, then the best pair of SPC charts to use would be….?
and X-bar chart and an R-chart
Six Sigma is a business improvement discipline whose fundamental view is based on what?
a performance-oriented approach to the business
Much of the Six Sigma methodology is used to identify and remove causes for what?
Process Variation
The distance between the Mean of a data set and the Point of Inflection on a Normal curve is called what?
the Standard Deviation
When we gather information for the Voice of the Business, we are primarily interested in information concerning what?
the profitability of the business
on a typical process map, what shape is used to depict a decision point requiring a Yes or No decision?
a Diamond
According to Cost of Poor Quality definitions, what kind of savings is Inventory Reduction?
Hard Savings
Customers make their decisions primarily based on 4 things:
- Features
- Integrity (of the seller)
- Delivery
- Expense
Which element of waste best describes “the unnecessary movement of materials and goods”?
Conveyance or Transportation
Order Production Cycle is defined as:
The amount of time between the receipt of an order and the receipt of payment
What could be defined as “a concise statement of the area of concern and why it is important that the issue be improved”?
Business Justification, aka Management Justification
The most important type of FMEA for a product BEFORE going into manufacturing is what?
Design FMEA
A major value derived from doing a SIPOC is that it provides insight into which inputs have the greatest affect on what?
on the CTQ outputs
Name a valuable tool used during the measure phase that can show material and information flow through out an entire process.
a Value Stream Map
Name a tool that is created for every project, based on the team’s collective opinions. It is a living document that gets updated whenever a parameter is changed.
an X-Y Diagram
What can a belt build in order to help him identify and evaluate risk factors for the subject process?
an FMEA
All of the data points that represent the total set of information of interest
the Population
Which distribution does not require a logarithmic base for probability calculations?
a Normal Distribution
the most frequently occurring value in a distribution of data
the Mode
If you wanted to model the data for the number of weaves in a section of carpet fabric, you would use which distribution?
a Poisson distribution
A correlation does not explain causation, but what can?
a Regression Analysis with a statistically valid mathematical equation
True or False: The Correlation Coefficient is always greater than the Regression Coefficient in a Simple Linear Regression
True
The calculation of column total times row total divided by Grand Total yields expected values from what?
a Contingency Table
Name some characteristics of a valid mathematical regression
- the sum of the residuals is zero
- the residuals when plotted follow a normal distrib.
- Most standardized residuals fall with +- 2 standard deviations
- a residual is equal to the difference between the observed and predicted values
What is commonly used to investigate and model the relationship between a response variable and one or more independent variables?
Analysis of Variance (ANOVA)
If the results from a hypothesis test are located in the “Region of Doubt” area, what can be concluded?
We reject the NULL Hypothesis
To be an effective LSS Practitioner, one must understand the difference between …..
Practical and Statistical significance
How do we establish a sample size that will allow the proper overlap of distributions?
Calculate 1 minus Beta
On which chart will one typically see a pattern from the graphed points such that conclusions can be drawn about the largest family of Variation?
a Multi-Vari Chart
What is defined as the difference between the observed and expected values for a given set of data?
Bias
the Greek letter sigma is used in mathematics to signify what?
Standard Deviation
Those people who have an interest in the output of a process are known as what?
Stakeholders
If a design has an alias, what does that imply?
that two factor effects are confused or confounded with each other
The theory of constraints focuses on what?
removing constraints and/or bottlenecks to increase throughput
All worthwhile management techniques require what?
monitoring, evaluation, review, and continuing improvement