Exam 2 Flashcards
In order for a problem to be solved correctly,
which of the following must occur first?
A. The problem must be defined. B. Relevant data must be gathered. C. The measurement system must be validated. D. The process must be mapped.
A. The problem must be defined.
Consider a process, which is assumed to follow a
normal distribution. If the nominal (target) is set
at the center of the distribution, and the
specification limits are set at ± 3σ from the
center, the Cpk is equal to:
A. 0.25
B. 1.00
C. 1.33
D. 1.67
B. 1.00
Comparing two popular tools, AIM analysis and
Fishbone analysis, what is most relevant:
A. Fishbone (Ishikawa) is more commonly used than AIM in the Analyze phase to organize potential causes to a problem B. Fishbone cannot replace AIM in the Define phase to visualize ideas and relation between problems C. AIM influences language and mind-sets in a higher extent than Fishbone does D. All of the above
D. All of the above
Compared to a two-level factorial experiment,
which of the following is an advantage of an
experiment using three levels?
A. Interaction effects can be estimated B. Curvature can be characterized C. The design can be augmented D. Efficiency is maximized in the experimental effort
B. Curvature can be characterized
The lack of Repeatability is defined as
A. the variation obtained when the same object is measured by the same operator with different measuring instruments B. the variation obtained when different operators are measuring the same object with the different measuring instruments C. the variation obtained when an operator is repeatedly measuring the same object with the same measuring instrument D. the variation obtained when different operators are measuring the same object with the same measuring instrument
C. the variation obtained when an operator
is repeatedly measuring the same object
with the same measuring instrument
What is the goal of Repeatability and Reproducibility studies (R&R-studies)?
A. to identify dominant sources of measurement variation B. to investigate whether the measurement system under study is usable for the intended task C. to identify control factors D. to identify interactions between CTQs
B. to investigate whether the measurement
system under study is usable for the
intended task
The most important aspect of functional
requirements is that they
A. describe a single, measurable performance B. describe how a product or service should operate C. are traceable to the voice of the customer D. provide upper and lower performance limits
C. are traceable to the voice of the
customer
The Pareto chart is a good tool for
A. identifying interactions between different
parameters
B. identifying whether a data set actually is
comprised of two sets of data with
different distributions
C. identifying the largest effects among a
large number of effects
D. finding trends in data sets
C. identifying the largest effects among a
large number of effects
For a normal distribution, two standard deviations
on each side of the mean would include what
percentage of the total population?
A. 47%
B. 68%
C. 95%
D. 99%
C. 95%
Correction, over-production, inventory, and
motion are all examples of
A. waste
B. 5S target areas
C. noise
D. value-added activities
A. waste
The main purpose of using robust design is
A. to maximize yield B. to remove noise factors C. to minimize the impact of noise factors on response variable D. all of the above
C. to minimize the impact of noise factors
on response variable
In the context of a 2K factorial design, if it is found
out that the interaction between factor A and
factor B is non-active, it means that the
interaction analysis proves that
A. factor A is independent of factor B (isn’t this correct to – partly) B. factor A is dependent on factor B and vice versa C. the effect of factor A on the response variable is dependent on the level of factor B D. none of the above
D. none of the above
Which of the following is a strategy to quantify
noise in a statistical experiment design?
A. Randomization
B. Replication
C. Blocking
D. Noise factorization
B. Replication
Which of these designs is saturated?
A. 27-3
B. 27-4
C. 27-5
D. None of the above
B. 27-4
When using X-bar chart, selecting sigma limit
lower than three (e.g. 2.5-sigma limit) will lead to
A. higher sensitivity of the chart B. lower ARL0 (average run length when the process is stable) C. higher type I error (alpha) D. all of the above
D. all of the above
Tools for cause and effect (hypothesis) testing are
listed below. Put them in their right place in the
table below:
A. Chi Square testing
B. Regression
C. Logistic Regression
D. Anova
A - D - C - B
Doing VA/NVA analysis from a VSM has the
objectives below. One is unfortunately wrong.
Which one is it?
A. Identify and eliminate the hidden costs that do not add value for the customer B. Reduce unnecessary process complexity C. Reduce process steps cycle time D. Reduce process lead time
C. Reduce process steps cycle time
The primary factor in the successful
implementation of six sigma is to have
A. the necessary resources
B. projects related to business measures
C. explicit customer requirements
D. a comprehensive training program
B. projects related to business measures
The control plan is one of major deliveries from a
Six Sigma project. It contains everything below
EXECPT:
A. Calculations and relation of R2 and R2
adj
B. Links between critical factors and big Y
C. SOP for preventive maintenance and
gauge control
D. Local control of process correction and
resources
A. Calculations and relation of R2 and R2
adj
R2 is recommended to be over a certain limit (e.g.
80%) in order to:
A. Ignore p-values above 5 %
B. Accept a model as a reasonable
explanation of the variation in the data
C. Claim that all residuals follow normal
distribution with zero mean
D. Eliminate the least significant factor from
the model
B. Accept a model as a reasonable
explanation of the variation in the data
A combination of both Shewhart control charts
and EMWA control charts is recommended
included in the control plan since:
A. The risk of false alarms decreases B. Amount of measurements decreases C. Number of both attribute and continuous measures monitored increases D. Sensitivity to both trend changes and special causes increase
D. Sensitivity to both trend changes and
special causes increase
When testing sample mean of bottle’s weight
using the following hypothesis, H0:μ0=500ml and
H1:μ0≥500ml, a p-value of 20% (n=40) is
obtained, this means that
A. The probability of the population’s mean
greater than 500ml (H1) is 1 in 5.
B. The probability of the population’s mean
equal to 500ml (H0) is 1 in 5.
C. Given that the population mean is 500ml
(H0), the probability of observing the
sample is 1 in 5
D. Given the observed sample, the
probability of the population mean is
500ml (H0) is 1 in 5.
C. Given that the population mean is 500ml
(H0), the probability of observing the
sample is 1 in 5
How much is the average change on flying time
(in seconds) as the Blade Length’s level is
changed from low level (-1) to high level (+1)?
A. 0.25
B. 0.50
C. 0.125
D. None of the above
A. 0.25
How much is the average interaction effect
between Blade Length and Paper Density?
A. -3
B. 0.15
C. -0.15
D. 3
C. -0.15
What is the best setting for the level of each
factor (BL=Blade Length, PD=Paper Density) as to
maximize the flying time?
A. low level for BL and low level for PD
B. low level for BL and high level for PD
C. high level for BL and low level for PD
D. high level for BL and high level for PD
C. high level for BL and low level for PD
In the context of x-bar chart, which of the following condition is not
assumed?
a. The distribution of the x-bar follows normal distribution
b. The distribution of the individual x follows normal distribution
c. The variance of the x-bar is never greater than the variance of the individual x
d. The average of the x-bar serves as an estimate of the process mean
b. The distribution of the individual x follows normal distribution
The process capability index Cpk
a. does not help analyzing variation
b. is always preferable to Cpl and Cpu
c. is good to be used when we suppose that the process is not centered on target
d. does not give an indication of dispersion and closeness to target
c. is good to be used when we suppose that the process is not centered on target
If the process is in statistical control, it implies that
a. The process has natural and unnatural variation
b. There is a statistically systematic variation in the process
c. There is no points falling outside the limits of the control chart
d. There is only random variation in the process
d. There is only random variation in the process
When testing the intercept and slope of a regression line, it is found out
that the p-values are much lower than 5%, it can be concluded that
a. The X-variable causes the Y-variable
b. The coefficient of determination is greater than 95%
c. The Y-variable is independent of the X-variable
d. None of the above
d. None of the above
With respect to linear regression analysis, which of the following is statement is false?
a. Linear regression analysis is sensitive to outlier
b. For the regression model to be valid, the residuals should be independently, identically, and normally distributed
c. One can use the regression model to predict both within and without the range of data in the X-variable
d. One can use the regression model to predict only within the range of data in the X-variable
c. One can use the regression model to predict both within and without the range of data in the X-variable
When the subgroup size of an x-bar chart is increased
a. Both the lower and the upper control limits will decrease
b. Both the lower and the upper control limits will increase
c. The upper control limit will increase, but the lower control limit will decrease
d. The upper control limit will decrease, but the lower
control limit will increase
d. The upper control limit will decrease, but the lower
control limit will increase
Which of the following statement is true in the case when the control limits of x-bar chart of subgroup size four, coincide with the specification limit, that is, UCL=USL and LCL=LSL?
A. Cp=Cpk=1
B. Cp1
D. None of the above
D. None of the above
If three-sigma control limits are used with one point data collected every day for the x-bar chart
a. There will be, on average, one false alarm every month
b. There will be, on average, one false alarm every six months
c. There will be, on average, one false alarm every year
d. There will be, on average, one false alarm every two years
c. There will be, on average, one false alarm every year
If four-sigma control limits are used instead of three-sigma, what will happen to the average run length when the process is stable (ARL0) and when the process is unstable (ARL1)
a. Both ARL0 and ARL1 will increase
b. ARL0 will increase, but ARL1 will decrease
c. ARL0 will decrease, but ARL1 will increase
d. Both ARL0 and ARL1 will decrease
a. Both ARL0 and ARL1 will increase
When there is one point falls outside either the lower or the upper limit of a control chart, it implies that
a. The process is stable
b. There is an assignable cause
c. The process is not capable
d. None of the above
d. None of the above
The purpose using potential-FMEA in a Six Sigma project is:
a. To create a list of underlying factor that might influence the variation in y
b. To visualize the status of the control of the most
influential factors in the underlying system for y
c. To sort out the most relevant factors influencing variation in y
d. To create a basic document as input for audit
b. To visualize the status of the control of the most influential factors in the underlying system for y
The control plan is one of major deliveries from a Six Sigma project. It contains everything below EXECPT:
a. Calculations and relation of R2 and R2
adj
b. Links between critical factors and big Y
c. SOP for preventive maintenance and gauge control
d. Local control of process correction and resources
a. Calculations and relation of R2 and R2
adj
In the Analysis of variance (ANOVA), the scope is to:
a. check the model assumptions
b. compare more than two treatments
c. verify that the error is normally distributed
d. show that a factor is significant
b. compare more than two treatments
The main difference between a control chart for variable and one
for attribute is:
a. We use the control chart for attributes when the process in control
b. We use the control chart for variables in combination with control limits
c. We use a control chart for attributes when we cannot
have quantitative measurements from the process
d. None of the above
c. We use a control chart for attributes when we cannot
have quantitative measurements from the process
DPMO stands for:
a. Define the Project, Measure and Observe
b. Dot Plot Multi Objective
c. Defects Per Million Opportunities
d. Defects, Problems, Measures, Optimization
c. Defects Per Million Opportunities
The DMAIC is:
a. Antithetic to the PDCA cycle
b. A framework for developing a six sigma project
c. A tool for obtaining six sigma quality
d. A tollgate
b. A framework for developing a six sigma project
Variation Mode and Effect Analysis is:
a. The equivalent of FMEA for large scale problems
b. A structured way to analyze variation of a key product
characteristic
c. A structured way to hide noise factors
d. None of the above
b. A structured way to analyze variation of a key product
characteristic
A statistical problem is:
a. A linear model with an error term
b. The conversion of a real problem into data and
methodology
c. The outcome of a six sigma project
d. The starting point of a six sigma project
b. The conversion of a real problem into data and methodology
The small ys and the CTQs in a Effective Scoping analysis:
a. Can be respectively considered as the input and the output of a process
b. Are always strongly correlated
c. The former are related to customer wants, the latter are
related to product features
d. Provide the direction for product improvement
c. The former are related to customer wants, the latter are related to product features
A robust design is characterized by:
a. Insensitiveness to noise factors
b. Absence of noise factors
c. Use of Design of Experiments
d. Low variation in performances
a. Insensitiveness to noise factors
Outliers:
a. are special causes of variation
b. need to be eliminated from a dataset
c. are highlighted only through probability plots
d. are opportunities for better understanding and
improvement
d. are opportunities for better understanding and
improvement
In a screening experiment with 7 factors the following result was obtained. What main effects are active and how will you label the red (square) dots in the normal probability plot from left to right?
a. C, A and E
b. F, A and G
c. C, D and G
d. F, D and E
d. F, D and E
The experiment, in question 22, has the aliasing pattern below. Based on the analysis above, what main effect might also contain a two-way interaction?
a. F = AG
b. A = CE
c. E = DF
d. G = CD
c. E = DF
Following the principles of sequential-DOE. What would you recommend for the next step of experimentation in this situation?
a. Full-factorial experiment
b. Folding
c. Replication
d. Center points
a. Full-factorial experiment
The interaction graph below was shown in a material alloying experiment. What is not a true observation:
a. A setting of Zr to high level will make Y robust to variations of C
b. C at high level will increase the sensitivity on Y of Mn
c. Y will increase when increasing Si if Zr is in high setting
d. Y is not dependent of the interaction between Si and C
b. C at high level will increase the sensitivity on Y of Mn
c. Y will increase when increasing Si if Zr is in high setting
By what means does Six Sigma contribute to economic, social or environmental sustainability?
a. Conducting a brainstorm
b. Designing an experiment
c. Reducing variation
d. Drawing a process map
c. Reducing variation
Even though the DMAIC phases are not serial, which one of the following is NOT an activity normally considered to be in the Analyze phase?
a. Study the causes of defects
b. Understand the sources of variation
c. Prioritize opportunities for future improvements
d. Develop a data collection plan
d. Develop a data collection plan
When you are running a BB project, what should you consider doing first of the following:
a. Evaluate the measuring system by performing an MSA
b. Define who needs the measurement result
c. Decide how the measurement result should be presented
d. Choose measuring device based on its accuracy
b. Define who needs the measurement result
In the project you are running, two operators handle the critical operation. They are supervising the operation and inspect the ready parts before shipment. Which of the following statements is FALSE?
a. The operation could be unstable
b. The operators’ judgment could be considered as a measurement system
c. There could be a problem with the repeatability and reproducibility of the inspection
d. It is required to have a standard describing how the inspection should be performed
d. It is required to have a standard describing how the inspection should be performed
What kind of experimental design is this?
(full factorial)
a. Saturated
b. Definitive Screening Design
c. Fractional factorial with resolution III, 2^(4-1)
d. None of the above
d. None of the above
Why is it important to have orthogonal columns in a DoE design table?
a. To make sure that the contrasts can be calculated independently.
b. To be able to calculate the maximum number of contrasts from a set number of degrees of freedoms.
c. It is the rows of the design matrix that should be orthogonal.
d. To allow for calculations of process variance.
a. To make sure that the contrasts can be calculated independently.
It was found out that the correlation between a response variable (Y) and a predictor (X) is low, that is, less than 0.1. It implies that
a. When the value of X is high, the value of Y is almost always high.
b. When the value of X is high, the value of Y is almost always low.
c. When the value of X is low, the value of X is sometimes high, but sometimes low.
d. When the value of X is low, the value of Y is almost always low.
c. When the value of X is low, the value of X is sometimes high, but sometimes low.
Tom is a Six Sigma black belt. He is very happy because his process, which is normally distributed, has a Cp of 2. Which of the following can be said about his process?
a. It is certain that the process is a six sigma process, having a defective rate of 3.4 ppm
b. It is possible that the process has more than 50% non-conforming items
c. It is possible that the control limits of the Xbar chart coincide with the specification limits, that is, UCL=USL and LCL=LSL
d. None of the above
b. It is possible that the process has more than 50% non-conforming items
Consider two explanatory variables (X1 and X2) which have substantial effects on a response variable (Y), in terms of statistical significance and effect size, when tested separately, that is, Y=f(X1) and Y=f(X2)). However, when they are tested together, Y=f(X1, X2), one of the explanatory variables becomes insignificant. Which of the following is very likely to be the cause?
a. The low correlation between X1 and X2
b. The high correlation between X1 and X2
c. The fact that X1 and X2 are orthogonal to each other
d. None of the above
b. The high correlation between X1 and X2
When using a control chart, one may have a false alarm. Which of the following statements about false alarm is WRONG?
a. A false alarm reflects randomness in a process
b. A false alarm can be set theoretically by changing the multiplier of the control limits (e.g. from three-sigma limits to four-sigma limits)
c. A false alarm is a signal when the process is under statistical control
d. None of the above
d. None of the above
When doing screening experiments the price for lower number of experiments is aliasing, which increases the risk of identifying the wrong significant factors and interactions. To support the evaluation of screening experiments, three guiding principles have been pointed out: Hierarchical Ordering Principle, Effect Sparsity Principle and Effect Heredity Principle. Which explanation below belongs to the Effect Sparsity Principle?
a. Lower order effects are more likely to be important than higher order effects.
b. The numbers of relatively important effects in a factorial experiment are small.
c. In order for an interaction to be significant, at least one of its parent factors should be significant.
d. Effects of the same order are equally likely to be important.
b. The numbers of relatively important effects in a factorial experiment are small.
What is the primary purpose with the control plan?
a. To guide what actions to take when process deviates in future
b. To follow-up on critical key performance indicators
c. To set future targets
d. To communicate project learning
a. To guide what actions to take when process deviates in future
In the supporting infrastructure for Six Sigma Black Belts developed by many organizations, some standard roles are identified. Which role is normally responsible for coordinating Six Sigma project selection?
a. Master Black Belt
b. Process owner
c. Deployment champion
d. Black Belt line manager
c. Deployment champion
Noise factors are the main target for both Six Sigma and DFSS. They are classified normally in five groups: aging, environmental conditions, usage, manufacturing variation and…?
a. Failure modes
b. Control factors
c. System interactions
d. Measurement noise
c. System interactions
The main purpose with this set-up of the Chalmers Six Sigma program is to:
a. train statistical engineering tools
b. solve a complex problem
c. understand that how -> what -> why
d. understand the application of the DMAIC phases
d. understand the application of the DMAIC phases
One of the following suggestions is false. Critical criteria for successful improvement projects are:
a. Learning and reflection
b. Management for improvement
c. Limited pressure for change
d. Problem analysis
c. Limited pressure for change
One of the following suggestions is false. Inner motivation is a crucial component for sustainable development. According to Deci & Ryan, the concept entails:
a. A shared vision
b. Autonomy
c. Personal development
d. Meaning
a. A shared vision
You are measuring a diameter on a machined axis using a
digital caliper. You get the value 20.02 mm. You have
performed a MSA and the result showed that the caliper is a good measurement system for this process. What is false?
a. The value 20.02 mm includes the true value and the measurement error
b. The axis has a true diameter of 20.02 mm
c. The Voice of the Customer states that the axis needs to have a diameter between 19 and 21 mm. I can mark the axis with a “OK”-sign.
d. If I measure more axes I can use the measurement result for process improvement analyses
b. The axis has a true diameter of 20.02 mm
‘Monitor y - manage x’ is a general driving force in Lean Six Sigma. What is most incorrect interpretation of that:
a. The control plan describes actions recommended when y deviates
b. Focus on x increase customer satisfaction and lower cost
c. Control charts can be used to visualize variation in x
d. Measurement systems for both x and y need to be qualified
a. The control plan describes actions recommended when y deviates
Empirical rule for scatter state that ±2s deviation from the
mean embrace X percent of the natural variation from any
distribution. X is:
a. 50-55%
b. 60-75%
c. 90-98%
d. 99-100%
c. 90-98%
In the analysis matrix (Figure 1) below what method replaces A?
a. logistic regression
b. t-test
c. linear regression
d. chi-square
d. chi-square
When paired t-test indicates a significant difference between two populations what characteristic is true?
a. The mean difference is large than zero
b. The mean difference significantly deviates from
a reference value
c. Zero is within the confidential band
d. Zero is outside the confidential band
d. Zero is outside the confidential band
An important diagnostics when analyzing data is to study
the residuals in order to find out if there is more information hidden in the data. What is NOT true about residuals:
a. Residuals are random and normal distributed with zero mean
b. Residuals show no trend versus time and order of measure
c. Residuals increase with the predicted model value
d. Residuals are used to calculate the standard error of the regression
c. Residuals increase with the predicted model value
What is a correct description of ‘a replicate’?
a. Multiple observations of the same experimental run (no adjustments of the settings, average of responses)
b. Minimizes within subgroup error
c. Gives information to predict measurement noise in the system
d. Duplication of a series of runs (takes error setting up equipment into account)
d. Duplication of a series of runs (takes error setting up equipment into account)
Within a control plan both XmR charts and EWMA charts are commonly proposed as new simultaneously used monitoring methods. Why?
a. XmR follow within sub-group variation, EWMA follow between sub-group variation
b. XmR is used for continuous data and EWMA for attribute data
c. XmR charts detects unusual large deviation from the mean, EWMA detects small shift/changes in a process
d. XmR is used for normal distributed data and EWMA is used for skewed data (one sided)
c. XmR charts detects unusual large deviation from the mean, EWMA detects small shift/changes in a process
Lean Six Sigma Black Belts that struggles with small
experimental budgets and many potentially significant
factors is guided by three principles for sequential DoE:
hierarchical ordering principle, effect scarcity principle, and effect heredity principle. What does latter one refer to (the effect heredity principle):
a. Lower order effects are more likely to be important than higher order effects.
b. In order for an interaction to be significant, at least one of its parent factors should be significant.
c. Effects of the same order are equally likely to be important.
d. The numbers of relatively important effects in a factorial experiment are small.
b. In order for an interaction to be significant, at least one of its parent factors should be significant.
Based on the parameter estimates results, what can be said about the relationship among the three variables (price, age of the clock, number of bidders)?
a. The age increases the number of bidders significantly
b. The age and the number of bidders are orthogonal
c. The price is significantly related to the age and the number of bidders
d. All of the above
c. The price is significantly related to the age and the number of bidders
Looking at the interaction plot, which of the following
statement is true?
a. The lower the price, the lower the effect of age on the number of bidders
b. The older the clock, the larger the effect of number of bidders change on the price
c. The larger the number of bidders, the smaller the effect of age change on the price
d. None of the above
b. The older the clock, the larger the effect of number of bidders change on the price
Which of the following statement is true for the Xbar chart?
a. The distribution of the x-bars follows normal distribution only if the distribution of the individual x-es follows normal distribution
b. The variance of the x-bars is always equal to or smaller than the variance of the individual x-es
c. The average of the x-bars will always be equal to or larger than the average of the individual xes
d. None of the above
b. The variance of the x-bars is always equal to or smaller than the variance of the individual x-es
Assuming that the process is stable, which of the following statement is true regarding the process capability?
a. The process is capable with a Cp value of more than 1 (Cp>1) because the control limits of the Xbar chart are narrower than the specification limits (USL and LSL)
b. The process is certainly not capable with a Cp value of less than 1
c. The Cp value of the process is smaller than the Cpk value because the process is off-centered
d. None of the above
b. The process is certainly not capable with a Cp value of less than 1
