Week 6 Self-Test Questions
Comparing the defect rate of two different processes is an example of a _________ experiment.
Binomial
Brad cannot remember the standard deviation of his process. The USL is 20, the LSL is 10, and the Cp is .833. What is the standard deviation?
Cp = (USL - LSL) / 6*StanDev, calculate backwards to get StanDev = 2
A golf ball manufacturer produces 200 golf balls per minute. Three types of defects can occur. The number of occurrences of each defect type over the same period is as follows: - Core material faulty: 1 - Cover is scratched: 1 - Numbering not readable: 2 Based on this data for a minute, what is the DPMO?
DPMO = (Defects* 1M)/(Total opportunity for defects) = (4*1,000,000)/(200*3) = 6666.7
A flag manufacturer produces 30 flags in a shift. Three types of defects can occur. The number of occurrences of each defect type over the same period is as follows: - Fretting: 3 - Color does not meet brightness standards: 2 - Spline issues: 5 Based on this data, what is the DPU?
DPU = number of defects/number of units DPU = (3+2+5)/30 DPU = 0.33/33%
True or False: Bill wants to collect data to form a subgroup over a period of time that is over the lunch break in an operation. This is a good idea.
False
True or False: Bob chose an IMR control chart to track the number of incorrect customer billings sent to customers per day. The IMPR control chart was the best choice that he could have made.
False
True or False: Bob wants to track the times of every delivery that he make on a control chart. He selects an XBarR chart. This is the most appropriate chart for this situation.
False
True or False: Jerry wants to collect data to form a subgroup over a longer period of time that cuts across multiple shifts in an operation. This is a best practice.
False
True or False: Jerry wants to collect data to form a subgroup over a longer period of time that includes the changing of machine type. This is a best practice.
False
True or False: Lower control limits are calculated based on customer requirements.
False
True or False: Marsha is using Poisson distribution for a dataset that is comprised of heights. Is this a best practice?
False
True or False: Marsha is working with a discrete data set, which she believes fits a normal distribution. Is this a sound idea?
False
True or False: Process capability is based on statistics with no consideration of the process/customer requirements.
False
True or False: SPC takes two forms: process assurance and process adherence.
False
True or False: There are no points outside of the lower control limit on a control chart, which means the process is under control.
False
True or False: There are no points outside of the upper control limit on a control chart, which means the process is under control.
False
True or False: Upper control limits are calculated based on customer requirements.
False
True or False: Bill wants to collect data to form a subgroup over a period of time where the product line is testing a new piece of machinery. This seem like a good idea.
False - Items within a subgroup should not be collected across a shift change or across any other process change that could be associated with a special cause
True or False: Jody is working with variable data. Based on this limited information, it is a best practice to use a u chart.
False, IMR Chart
True or False: When sampling with subgroups of different sizes, it may be appropriate to use an IMR chart.
False, IMR charts don't use subgroups
True or False: To create a control chart of defects over a constant size of area of opportunity (such as X defects/100), a p chart should be used.
False, NP chart
True or False: Bob wants to create a control chart for missed opportunities in a sample of 100 customer interactions per day so he decides to use a IMR chart. In this situation, Bob selected the best chart.
False, NP chart would be most effective here: non-conforming items over a constant area of opportunity
True or False: Marsha is using Poisson distribution for a dataset that is comprised of widget weights. Is this a best practice?
False, Poisson is a discrete distribution
True or False: Bob has calculated Ppk, Pp, Cp, and Cpk; however, he has forgotten that his data are non-normal. Therefore, only the Pp is a valid calculation.
False, Ppk and Pp
True or False: Janet has data on the outcome of coin tosses and wants to use statistical tools tailored to Weibull distribution. This is a best practice.
False, Weibull is a continuous probability distribution (infinite outcomes)
True or False: A local food truck company wants to monitor, via a control chart, the number of trucks that are out of service (i.e., defective) each day. The control chart the manager selects for this task is an IMR chart. This is the most appropriate control chart.
False, likely need NP or P Chart
True or False: Bob is dealing with discrete data and wants to use a statistical tool for data that fits a normal probability distribution. Is this a best practice?
False, normal is continuous distribution
True or False: Julien is trying to determine his variable type in order to select the proper statistical tests. He is measuring the grade level of a student. He believes this to be a continuous variable. He is correct.
False, this is a discrete vairable
What is the PPM if the DPU is .2? Question options: a) 100,000 b) 200,000 c) 500,000 d) none of the above
PPM = DPU x 1M = 0.2 x 1M = 200,000
A process has a USL of 10, a LSL of 8, and a standard deviation of 2. What is the Pp?
Pp = (USL - LSL)/6*standev = (10-8)/6*2 = 0.1666
A process at a large biotech manufacturing facility goes through a four-step process with the following yields: .9, .95, .8, and .99. What is the RTY?
RTY = 0.9 * 0.95 * 0.8 * 0.99 = 0.677
True or False: Brad is reviewing his control chart and discovers that he has 15 points in a row that are above the mean, but under the UCL. This would be an example of _________ that would need to be further explored.
Special Cause Variation
A process has a DPU of 15%. What is the TPY? a) 80% b) 85% c) 90% d) none of the above
TPY = number of accepted units/number of units 100% - 15% = 85% TPY
A process at a large biotech manufacturing facility goes through a four-step process with the following yields: .9, .95, .8, and .99. What is the total yield of this process for a shift if 1,000 units are inputted?
TPY: 1000*(0.9) = 900*(0.95) = 855*(0.8) = 684*(0.99) = 677
True or False: A C-chart is used with counts of attribute data.
True
True or False: A common way to identify the appropriate probability distribution is to use statistical software.
True
True or False: A process consisting of only common cause variation is a process that is in control.
True
True or False: Bob notices that he has only point above the upper control limit. Based on the common rules to detect special cause, this process is potentially out of control.
True
True or False: Capability is assessed by comparing process variation to customer expectations.
True
True or False: Julien is trying to determine his variable type in order to select the proper statistical tests. He is measuring the time it takes him to get to work. He believes this to be a continuous variable. He is correct.
True
True or False: Process capability is obtained by comparing the process average (mean) to the customer specifications.
True
True or False: The central limit theorem states that, regardless of the population shape, the sampling distribution of the mean approaches a normal distribution (bell curve) if the sample size is large enough.
True
True or False: When sampling with subgroups of different sizes, it may be appropriate to use a u chart.
True
True or False: When working on a process improvement, it is best to jointly consider the "voice of the process" and the "voice of the customer."
True
True or False: X-bar is a term to describe the mean of a subgroup.
True
True or False: Knowing only the mean and standard deviation, it is possible to compute the lower control limit on an I chart.
True, UCL/LCL = overall average +/- 3 standevs
True or False: When reviewing data pertaining to customer wait times that fit an exponential distribution, without knowing more information, it would be expected that there are fewer large values and more small values.
True, this is the definition of exponential distributions
A scatter diagram is used to study and identify the possible relationship between the changes observed in two different sets of __________.
Variables
A business tracks the cost for every audit they conduct. They wish to put this on a control chart. Which of the following control charts is the most appropriate? a) IMR b) XbarX c) c chart d) h range chart
a) IMR (subgroup size = 1)
Of the options below, which Ppk is the most ideal? a) 0.23 b) 1.2 c) 1.45 d) None of the above
c) 1.45
Jerry is running a large factory that produces 100 units per hour. He wants to track the weight of the product by collecting five concurrent samples per hour each day. Which of the following is most appropriate? a) XBarR b) np chart c) IMR d) p chart
a) XBarR
Julien is trying to determine his variable type in order to select the proper statistical tests. The data he collected deal with the number of defects on a production line. What type of variable should he use? a) discrete b) compounding c) pounds d) continuous
a) discrete
A process has a Ppl of .5 and a Ppu of 1. What is the Ppk? a) 0 b) 0.5 c) 1 d) none of the above
b) 0.5, Ppk is the smallest value of either Ppu or Ppl
Which of the following is an input into the Control Phase? a) additional Xs that still need to be tested b) an improved process that is stable and predictable and meets customer requirements c) a Y = f (x) - transfer function that still cannot be explained d) a process that has not been implemented
b) an improved process that is stable and predictable and meets customer requirements
Bob's boss is excited to see if a new process is under control four months after a Six Sigma project. However, no information seems to be available. It seems that _____________ should have been part of the project's SPC/control plan. a) process capability studies b) control charts c) histograms d) cpk and ppk measures
b) control charts
Tossing a coin 100 times and determining how many times heads will occur is an example of a _________ experiment.
bi-nomial
A 60% process yield equates to which of the following? a) 60,000 DPMO b) 600,000 DPMO c) 400,000 DPMO d) 40,000 DPMO
c) 400,000 DPMO 60% = 40% defective, multiply by a million
Bob has calculated Ppk, Pp, Cp, and Cpk; however, he has forgotten that his data are non-normal. Which of the following is true? a) Only Ppk, Pp, and Cp are valid calculations. b) Only Cp and Pp are valid calculations. c) Only Ppk and Pp are valid calculations. d) All are not valid.
c) Only Ppk and Pp are valid calculations.
Suppose there are data on the number of dents on various cars after a hail storm. Based on this information, which distribution is likely the best fit for this data? a) normal b) Gaussian c) Poisson d) hypoallergenic
c) Poisson - cracks in a windshield, no upper bound on count
Which of the following would be an example of special cause variation of the on-time delivery policy for a food truck? a) non-extreme traffic affecting a delivery b) rain in April that does not cause major impact c) a blizzard strikes in a warmer part of the country that causes major delays d) Normal stops for gas
c) a blizzard strikes in a warmer part of the country that causes major delays
Jody is working with attribute data. Based on this limited information, which of the following is the best control chart type for this situation? a) XbarS b) RangeBar2 c) c chart d) XbarR
c) c chart - only option that is used for attribute data
Brad is trying to determine the proper distribution for the time customers are waiting before an appointment. In the absence of other information, a good bet may be that his data fit a __________ distribution. a) Poisson b) gamma c) exponential d) binomial
c) exponential - used to model the time elapsed between events
Bob is dealing with discrete data and the samples are without replacement. Based on this information, what is the likely probability distribution? a) Weibull b) binomial c) hypergeometric d) normal
c) hypergeometric
Bob wants to create a control chart for the number of incorrect customer billings sent to customers per day. Which of the following is the best control chart to use? a) IMR b) XbarR c) p chart d) UCL chart
c) p chart
Before Joe selects the proper statistical tools to use in the Analyze Phase, it is best that he determines the right ___________ during the Measure Phase. a) Poisson values b) P-values c) probability distribution d) R^2
c) probability distribution
Brad is reviewing his control chart and discovers that he has 15 points in a row that are above the mean, but under the UCL. This would be an example of _________ that would need to be further explored. a) common cause b) standard deviations c) special cause variation d) triple threat customization
c) special cause variation
An aerospace manufacturer wants to monitor the scratches (defects) on a gear. Employees record the number of scratches on each gear, and each production subgroup has a different number of gears. Which control chart is most appropriate? a) IMR b) Zstd c) u chart d) Rf chart
c) u chart - defects per unit, different subgroup sizes
Marsha has a data set that fits the normal distribution. The mean is 10 and the standard deviation is 1. How much of the data can be expected to fall within two standard deviations of the mean? a) ~24% b) ~68% c) ~95% d) ~99%
c) ~95% (2 standevs away from mean is 95.44%)
A business tracks the days it takes to deliver every product it sells within a 20-mile radius. They wish to put this on a control chart. Which of the following control charts is the most appropriate? a) u chart b) XbarX c) c chart d) IMR
d) IMR, subgroup size is the same
Bob is dealing with a dataset comprised of time to failure data on car engines. Based on this information, which of the following distributions potentially offers the best fit? a) Betronics b) binomial c) hypergeometric d) Weibull
d) Weibull
Jill is managing a large call center that takes 2,000 calls per hour. She wants to track the wait times of callers by collecting three concurrent samples per hour each day. Which of the following is most appropriate? a) p chart b) np chart c) IMR d) XBarR
d) XBarR - variable subgroup size
Process capability analysis ________________. a) is a scientific and systematic procedure that uses control charts b) answers the question: "Is this process capable of meeting the customer's requirements?" c) determines if there has been some type of change in the process d) all of the above
d) all of the above
Marsha is working with a discrete data set. Which of the distribution choices below is likely to fit her data? a) normal or exponential b) Weibull or normal c) binomial or Weibull d) binomial or hypergeometric
d) binomial or hypergeometric
Julien is trying to determine his variable type in order to select the proper statistical tests. He is measuring the weight of a part. What type of variable is this? a) discrete b) compounding c) lbs. d) continuous
d) continuous - quantities such as pressure, height, mass, weight, density, volume, temperature, and distance are examples of continuous variables
What tool is the best one to determine the "voice of the process"? a) Cpk b) Cp c) Venn diagram d) control chart
d) control chart
Brad is trying to determine the proper distribution for a set of reliability data. In the absence of other information, a good bet may be that this set of data fits a ______________ distribution. a) Poisson b) gamma c) binomial d) exponential
d) exponential - exponential distribution is widely used in the field of reliability. Reliability deals with the amount of time a product lasts. Other uses include probability of arrival times, distance or space between occurrences of events of interest, and waiting times.
Jim wants to put counts of defects over a constant size of data on a control chart. Which of the following choices is the best control chart to use? a) normal X chart b) IMR c) Xsr d) np chart
d) np chart - tracks defective units over a constant area
The intent of control charts is to identify __________ variation. a) total process b) significant c) common d) special cause
d) special cause
Marsha has a data set that fits the normal distribution. The mean is 10 and the standard deviation is 1. How much of the data can be expected to fall within one standard deviation of the mean?
~68% (based on empirical rule)