Lean Six Sigma Green Belt - Measure Module

The short-term Z-score is 4.0 after a process improvement effort. What will the long-term Z-score be? 3.5 2.5 5.5 4.5

2.5 Long-term Z = Short-term Z - 1.5.

In a box plot, the interquartile range (IQR) contains _________ of data studied. 25% 90% 50% 75%

50% The Interquartile range (IQR) = Upper quartile (75%) - Lower quartile (25%) = 50%. It represents the size of the box in terms of data values.

For Cpk assessment, the process should be _________ and data must be _______. A) stable; normal B) normal; stable C) stable; long-term D) short-term; normal

A Cpk is used to predict the performance of a process. It should be assessed based on a stable process and normal data.

Data is important for gaining knowledge about: *Customer needs. *Problems. *Processes. *All of the above.

All of the above. Data is used to understand the customers' needs, real problems, root causes of problems, and processes and how to control processes.

Value-added flow analysis helps with identifying: *Value-added activities. *Non-value added activities that are necessary. *Non-value-added activities that are not necessary (waste/muda). *All of the above.

All of the above. The purpose of value-added flow analysis is to figure out value-added and non-value-added activities (necessary and unnecessary) to improve the "value" of products/services from the customers' perspective.

The consistency in the inspection/review process can be measured by: *Attribute MSA. *Variable MSA. *Process capability analysis. *Analysis of variance.

Attribute MSA Attribute agreement analysis/attribute MSA provides assessment about how inspectors/reviewers are conforming to known criteria.

A p-chart is used for _________ and a u-chart is for _______. A) defect data; defective data B) defective data; defect data C) continuous data; attribute data D) fixed samples; variable samples

B A p-chart is used to plot the percent of defective items, whereas a u-chart is used to plot the number of defects per unit. Sample size is not fixed for both.

The problem in accuracy, linearity, and/or stability result in the average of the data being below or above the true value. The distance between the average and true value is known as: *Repeatability. *Bias. *A defect. *The Z-score.

Bias Bias is the offset from the true value. Bias is defined as the deviation of the measured average value from the actual value. Calibration is a way to minimize and control bias within acceptable limits.

At a customer service center, total incoming calls = 10,000 during a month. Total dropped calls before they were answered = 1,333. What distribution is suitable for process capability estimation? *Binomial *Normal *Poisson *Skewed

Binomial Dropped calls are defective data. Binomial distribution is used for defective (binary type of) data.

What term is used for the ability of a measurement system to detect small changes or granularity in the characteristics of a process? *Resolution *Discrimination *Both of the above *None of the above

Both of the above Resolution refers to how detailed the information is. Discrimination is the smallest detectable increment between two measured values. If discrimination of the measurement system is so poor, it sees virtually everything the same.

If Cpk upper (Cpu) = 2.0 and Cpk lower (Cpl) = 0.83, what can you conclude? A) The process is stable. B) The reported CPk is 2.0. C) The process is not stable because Cpk is less than 1.33. D) There is not enough information.

C The reported Cpk is the minimum of two values (Cpu, Cpl). Cpl = 0.83, so if Cpk = 0.83 < 1.33, we can conclude that the process is not stable.

If one is checking measurements against a reference/standard, she is doing a(n) ___________ check. *repeatability *stability *accuracy *linearity

Checking measurements against a reference/standard/master is an accuracy check.

The ratio and interval scale of data are linked to the___________ type of data. *continuous *binary *nominal *ordinal

Continuous Continuous data is a type of quantitative or numerical data, which can use a ratio scale as well as an interval scale. Other options are related to the type of attribute or qualitative data.

In measurement system analysis, precision can be divided into ___________ and _____________. A) repeatability; accuracy B) reproducibility; stability C) accuracy; bias D) repeatability; reproducibility

D Precision means getting the same values for the characteristics of a process every time a measurement is made using the same device/equipment. Repeatability is observed when one operator obtains measurements, whereas reproducibility is observed when more than one operator obtain measurements.

Which of the following is a technique for organizing data into different groups for analysis? *Data sampling *Data monitoring *Data stratification *Data controlling

Data Stratification. Data stratification is the process of partitioning data into distinct or nonoverlapping groups based on a predetermined set of criteria (e.g., age, shift, location, gender). Stratification improves the precision of results and is an efficient sampling technique.

A decision-making activity is represented by a _____________ symbol in the detailed process map. *circle *box *arrow *diamond

Diamond Circle = Operation Box = Inspection Arrow = Movement Diamond = Decision

The _________ of individual values can be a useful alternative to the histogram. *scatter plot *Pareto chart *time series plot *dot plot

Dot plot The dot plot can be a useful alternative to the histogram, especially if you want to see individual values or you want to brush the data. Brushing is the process of interactively highlighting data values on a graph.

T/F Attribute agreement analysis is different from measurement system analysis.

False Attribute agreement analysis is the measurement system analysis for attribute data (e.g., pass/fail, accept/reject).

T/F Long-term process capability is subject to special cause variation only.

False Long-term process capability is subject to both common and special causes of variation.

T/F Measurement system analysis is used for continuous data only.

False Measurement system analysis can be used for both continuous and attribute data.

T/F Value stream mapping (VSM) does not include suppliers and customers.

False Value stream mapping (VSM) includes suppliers, high-level process steps, information flow, material flow, and customers.

T/F The 5 Whys approach is a problem-solving technique.

False. 5 Whys is not a problem-solving tool but an effective root cause analysis technique.

If % tolerance GR&R ≤ 10%, then the measurement system (MS) is: *Poor. *Acceptable. *Marginal. *Ideal.

Ideal If the % tolerance GR&R ≤ 10%, then the MS is ideal. If the % tolerance GR&R = 10%-20%, then the MS is acceptable. If the % tolerance GR&R = 20%-30%, then the MS is marginal. If the % tolerance GR&R > 30%, then the MS is poor.

Process maps help in characterizing the functional relationships of various: *Inputs and noise factors. *Outputs and requirements. *Inputs and outputs. *Inputs and business needs.

Inputs and outputs Process mapping is a graphical team effort to map out all activities in a process to study how inputs may affects outputs.

The x-axis of a histogram contains the ______________scale of data. *interval *nominal *ratio *ordinal

Interval A histogram is an special type of bar graph. It displays data that have been summarized into intervals on its x-axis.

Lead time for a delivery process is the time elapsed between when a product is ordered and when the product: *Has started to produce. *Is shipped. *Is delivered. *Is ready for delivery.

Is delivered Lead time is the time interval between the initiation and completion of a process.

In attribute agreement analysis, the _________ is used to measure inter/intra-rater agreement. *t-statistic *z-statistic *kappa statistic *F-statistic

Kappa statistic The kappa statistic (or kappa coefficient) is the most commonly used statistic for measuring inter-rater reliability for qualitative/attribute data.

A distribution that best fits the process data is selected based on p-value. What distribution should be selected for process capability analysis with a confidence level of 95%? 1. Weibull distribution, p-value < 0.003 2. Lognormal distribution, p-value = 0.075 3. Exponential distribution, p-value < 0.005 4. Normal distribution, p-value < 0.05 *Weibull *Lognormal *Exponential *Normal

Lognormal For lognormal distribution, p-value = 0.075 > 0.05 (α-risk). α-risk = 1 - confidence level = 1 - 0.95 = 0.05.

__________ distribution is the non-normal distribution for continuous data. *Binomial *Poisson *Lognormal *Gaussian

Lognormal Lognormal distribution is non-normal distribution for continuous type of data. Gaussian distribution is also known as normal distribution. Binomial and Poisson distributions are used for attribute data.

What method is used to know if the data is accurate and precise? *DMAIC *Measurement system analysis (MSA) *Process capability analysis *5S

MSA Data is considered accurate and precise if the measurement system is accurate and precise. MSA provides assessment on the measurement system accuracy and precision.

The variation in the measured data set due to the way measurements are being taken is known as _______________ error. *measurement *sampling *estimation *human

Measurement Measurement error is the total effect of all sources of measurement variability that cause a measured value to deviate from the true value.

The preferred measure of central tendency of skewed continuous data or ordinal data is: *Median. *Mode. *Mean. *Trimmed mean.

Median Median is the physical center of a data set and not affected by extreme data values.

An uncontrolled or difficult-to-control variable that is identified in a fishbone diagram is known as: *Controllable. *Fixable. *Procedural. *Noise.

Noise Controllable variables are adjusted or controlled while the process is running. Procedural variables are established under certain standards or compliance requirements to run a process. Noise refers to variables that are unknown, uncontrollable, or too expensive or too difficult to control.

Which of these is not continuous data? 25 km distance Number of students 5 cm long 3 liters of milk

Number of students The number of students is discrete data (countable data).

The Cpk will be negative if the mean of a process output is: *Away from the target. *Outside the customer specification limits. *Equal to one of the customer specification limits. *Inside the customer specification limits.

Outside the customer specification limits. Cpk = min.{(mean - LSL)/3σ, (USL - mean)/3σ} LSL: Lower Spec Limit. USL: Upper Spec Limit. If mean < LSL or mean > USL, Cpk will be negative.

Process capability is assessed by comparing the Voice of the ___________ to the Voice of the ___________. Process; Business Business; Customer Procedure; Customer Process; Customer

Process; Customer Process capability is the inherent ability of a process to meet the customer's expectations. In other words, it is how much the VOP can meet the VOC.

The characteristics of the appropriate measure are sufficient, relevant, _______________, and contextual. *representative *usable *suitable *appropriate

Representative Sufficient means that the measures are available to be measured regularly. The measures are relevant to the process, so they can help to understand and isolate the problems. They are representative of the process across shifts and people, so the variation can be detected across shifts and people. They are contextual, meaning it is necessary to collect information with other relevant information that might explain process variability.

What should you do first if the high variation in data is coming from the measurement system, assuming the equipment has no issue? *Send the equipment for calibration. *Retrain the operators. *Repeat the MSA. *Improve the production process.

Retrain the operators The primary contributors to the high variation associated with the measurement system are repeatability and reproducibility. The equipment is not an issue if it working properly. We should retrain the operators about the procedure and standards for better understanding.

For a normal distribution, the mean, median, and mode are same. If the mean is greater than the median, the distribution is: *Left skewed. *Right skewed. *Symmetrical. *Bimodal.

Right skewed The mean is greater than the median due to extremely high data values. It makes the distribution tail longer on the right side, which is known as right-skewed distribution.

Which of the following is a high-level process map? *Value stream map (VSM) *SIPOC *Fishbone diagram *Swim lane process map

SIPOC SIPOC (Supplier-Input-Process-Output-Customer) provides the mapping of the high-level activities of a process.

What is the technique/process used in statistical analysis to get a small group of data to represent the whole population of data? *Stratification *Sampling *Segmentation *Sorting

Sampling Sampling is the basis for statistical inference; if it is done appropriately and accurately, then the inference is that what we see in the sample is representative of the population.

Which of the following is not a focus of the FMEA approach? *Improving the quality *Reducing product development time *Solving problems *Preventing problems from occurring

Solving problems FMEA helps to focus on continuous problem prevention, not problem solving. FMEA helps to improve quality, reliability, and safety; reduce development time and cost; and increase the level of customer satisfaction.

The X-Y matrix is a prioritization tool for the potential Xs based on the: *Management decision. *Team's collective opinions. *Managers' instructions. *Company policies.

Team's collective opinions. The X-Y matrix is a team-based prioritization tool for the potential Xs. It is created for every project and updated whenever a parameter is changed.

A data set is considered to be normal with a 95% confidence level if: *The p-value > 0.05. *The p-value ≤ 0.05. *Z = 6.0. *The p-value > 0.005.

The p-value > 0.05 Confidence level = 95%, so risk = 5%, or 0.05. If the p-value > 0.05, then data follow a normal distribution.

T/F A spaghetti diagram/plot is used to trace the movement of products, papers/documents, and people.

True A spaghetti diagram is one of the Lean tools. It is used for determining the distance traveled by people and/or materials and for reducing the wastes related to conveyance/transportation, motion, and waiting.

T/F An operational definition is a clear, concise description of a parameter or variable of a process.

True An operational definition helps to ensure common, consistent, and error-free interpretation of results or outcomes of a process.

T/F If the data distribution is skewed, range is used to estimate variability of data.

True Either range or interquartile range is used when the data distribution is skewed or has outliers.

T/F In the Six Sigma approach, the current process capability is the key output of the Measure Phase.

True In the Measure Phase, the practical problem is converted into a statistical problem in terms of the current process capability to show what the probability is of our process producing a defect-free output.

T/F Measurement system analysis also includes the environment.

True Measurement system analysis includes the operator, reference/standard, procedure, equipment, and the environment.

T/F A data collection plan is used mostly to gather passive data.

True Passive data collection means that data is collected without messing with the process.

T/F A process is not stable if process data show any special cause variation.

True Stable processes do not have any special cause variation; they only have common cause variation.

T/F The purpose of the Measure Phase is to identify the gap in process capability in order to meet a target or the Voice of the Customer (VOC).

True The purpose of this phase is to establish the baseline performance of the current process, which is used to assess the improvement in the Improve Phase.

T/F The following customer service time data are collected at a bank branch for 20 subgroups, for which there are 6 customers per subgroup. Overall mean = 2.5 min., and average range (R-bar) = 1.0 min. Upper and lower specification limits are 1.5 min and 4.0 min. The probability of customers served below 1.5 min is greater than that of customers served above 4.0 min. True False

True Zl = (LSL - mean)/std. dev. Zu = (mean - USL)/std. dev. std. dev. = R-bar/d2 = 1.0/2.534 = 0.4 min (approx.) (d2 = 2.534 for subgroup size of 6) Zl = (1.5 - 2.5)/0.4 = 1.0/0.4 = -2.5, Probability = 0.621% Zu = (4 -2.5)/0.4 = 1.5/0.4 = 3.75, Probability = 0.009% Note: Use the IASSC reference document for the d2 value (Control charts - Table of constants) and for probability from the Z-table.

What does VSM stand for? *Value system mapping *Value system management *Value-added system mapping *Value stream mapping

Value stream mapping

Attribute process capability is always assessed based on long-term data, as it requires ____________ samples to get a good estimate statistically. *too few *few *so many *all

so many As compared with continuous data, attribute data are qualitative and subjective data. We need more samples for any estimation to be statistically sound.