Free CSSGB Practice Test Questions and Answers (2026)

The foundational premise of Lean Enterprise, derived from the Toyota Production System, is that any process or work activity contains waste (Muda). Lean philosophy posits that waste is any step or action that consumes resources but does not add value from the customer's perspective. The primary goal of Lean is to systematically identify and eliminate this inherent waste through continuous improvement, thereby maximizing value and optimizing workflow. Therefore, wherever work is being done, it is assumed that waste is also being generated.

A. While work involves spending money, this is a general business fact, not the specific operational premise that drives Lean methodology's focus on process improvement.
C. This is the desired outcome or goal of a process, not the assumed reality. Lean operates on the premise that much of the work being done is non-value-added (waste).
D. This is the objective and activity of applying Lean principles. The premise is that waste exists to be eliminated, not that it is already being eliminated.

1. Womack

J. P.

& Jones

D. T. (2003). Lean thinking: Banish waste and create wealth in your corporation. Simon and Schuster. In Chapter 1

the authors define Lean thinking starting with specifying value

which inherently separates value-added activities from non-value-added activities (waste). The premise is that waste exists in all processes (p. 15).

2. Shah

R.

& Ward

P. T. (2007). Defining and developing measures of lean production. Journal of Operations Management

25(4)

785-805. The paper identifies "waste elimination" as a core construct of Lean production. The practice is predicated on the assumption that waste is present in processes (Section 3.1

p. 791). https://doi.org/10.1016/j.jom.2007.01.001

3. Purdue University. (n.d.). Lean Principles. Purdue Manufacturing Extension Partnership. The first principle listed is "Identify Value

" which involves defining value from the customer's perspective to recognize what is waste. The second principle

"Map the Value Stream

" is used to "identify and document all the activities... both value-added and non-value-added (waste)." This confirms the premise that waste is present to be identified.

4. Hines

P.

Holweg

M.

& Rich

N. (2004). Learning to evolve: A review of contemporary lean thinking. International Journal of Operations & Production Management

24(10)

994-1011. This review emphasizes that the "starting point for lean thinking is the recognition of waste" (p. 997)

reinforcing that the presence of waste is the fundamental premise. https://doi.org/10.1108/01443570410558049

The statement is true. In the practice of statistical hypothesis testing, particularly within the Six Sigma framework, certain conventional risk levels are widely adopted. An Alpha (α) risk of 0.05 is the most common standard for the significance level, corresponding to a 95% confidence level. This is the risk of a Type I error (false positive). A Beta (β) risk of 0.10 is also a very common and desirable target for Type II error risk (false negative). This Beta level corresponds to a statistical power of 90% (1-β), which represents a high probability of correctly detecting an effect when one truly exists. Therefore, these values are considered standard practice for a robust statistical test.

B. False: This option is incorrect because α = 0.05 and β = 0.10 are indeed the most frequently cited and utilized conventional values for setting Type I and Type II error risks in applied statistics and quality improvement methodologies like Six Sigma.

1. Montgomery

D. C. (2013). Introduction to Statistical Quality Control (7th ed.). Wiley.

Page 139

Section 4-2.2: "The choice of α = 0.05 is very common

although other values are also used."

Page 147

Section 4-3.2: In an example for determining sample size

the text uses a desired power of 0.90

which directly implies a Beta (β) risk of 0.10 (1 - 0.90). This combination is used to illustrate standard practice.

2. NIST/SEMATECH. (n.d.). e-Handbook of Statistical Methods. National Institute of Standards and Technology.

Section 1.3.3.1

Significance Levels: "The most frequently used value for alpha is 0.05."

Section 7.2.2.2

Sample sizes required: The examples provided for calculating sample size frequently use a power of 90% (β = 0.10) in conjunction with a significance level of α = 0.05

reinforcing this combination as a common standard.

3. The Pennsylvania State University. (n.d.). STAT 501: Regression Methods. Penn State Eberly College of Science.

Lesson 1: Simple Linear Regression

Section 1.10: The courseware notes

"The choice of the significance level

α

is admittedly arbitrary. However

the choice of α = 0.05 is the most common." The concept of power (1-β) is also discussed

with 80% or 90% being typical goals.

The payback period is the time required for the cumulative savings from a project to equal the total initial investment. First, calculate the total initial investment by summing the project investment and training costs: $47,500 + $6,500 = $54,000. Next, determine the time required to recover this investment once savings begin: $54,000 / $3,500 per month ≈ 15.4 months. The problem states that savings begin after 4 months. This initial delay must be added to the recovery time. Therefore, the total payback period is the sum of the delay and the recovery period: 4 months + 15.4 months = 19.4 months.

A. 9.7: This figure does not correspond to a correct calculation of the payback period using the provided data.
B. 15.4: This represents only the time to recover the investment after savings have started, but it incorrectly omits the initial 4-month delay period.
D. 23.7: This figure is not derivable from a standard payback period calculation and likely results from a miscalculation or misinterpretation of the inputs.

1. ASQ Certified Six Sigma Green Belt Body of Knowledge (BoK): The BoK, under Section I.F.3 (Business case and project charter), requires understanding financial measures to justify projects. The payback period is a fundamental metric for this purpose, evaluating the time to recoup the initial investment. The calculation method used is standard financial practice for this metric.

2. Park, C. S. (2019). Contemporary Engineering Economics (7th ed.). Pearson. In Chapter 7, "Payback Period Analysis," the payback period is defined as the time required for the cumulative net cash flow to equal the initial investment (p. 298). The calculation involves dividing the initial cost by the periodic savings, and any initial period with no cash flow is added to this time.

3. Thuesen, G. J., & Fabrycky, W. J. (2001). Engineering Economy (9th ed.). Prentice Hall. Chapter 5, Section 5.3, "The Payback Period," describes the method as calculating the period of time required for the profit or savings of an investment to equal its cost. The formula is given as Payback Period = Initial Cost / Uniform Annual Benefit (p. 178). The example demonstrates that the clock starts at time zero, including any non-productive periods.

Kanban is a visual signaling system used in Lean methodologies to manage workflow and trigger actions. While its most famous application is in controlling inventory and production flow (e.g., signaling the need to produce or move parts), the underlying principle is to use a visual cue (like a card) to signal a need. In the context of a Visual Factory, this principle is extended to other processes. A card placed in an area that requires cleaning or organization acts as a "Kanban" signal to the relevant team, making the need for action immediately visible and prompting a response.

B. Kaizen: This is a philosophy of continuous improvement involving everyone, not a specific type of card used to signal cleaning needs.
C. Poke-Yoke: This is a mistake-proofing or error-proofing method designed into a process to prevent defects from occurring.
D. WhoSai: This is not a recognized term or tool within the Lean Six Sigma or Visual Factory body of knowledge.

1. Liker, J. K., & Meier, D. (2006). The Toyota Way Fieldbook. McGraw-Hill. In Chapter 10, "Build a Lean Learning Enterprise," the authors discuss the broad application of visual controls to make any abnormality or need for action immediately apparent, which is the core function of a Kanban signal.

2. Dennis, P. (2016). Lean Production Simplified: A Plain-Language Guide to the World's Most Powerful Production System (3rd ed.). CRC Press. In Chapter 5, "Kanban," Dennis defines Kanban as a "sign" or "signal" that triggers replenishment or action, emphasizing its role as a fundamental visual control tool.

3. Wilson, L. (2010). How to Implement Lean Manufacturing. McGraw-Hill. In Chapter 10, "The Visual Workplace," the text explains that visual signals are used to highlight abnormal conditions. A card indicating a need for cleaning is a type of visual signal that functions on the same principle as a production Kanban.

4. MIT OpenCourseWare. (2005). 16.852J / ESD.60J, Integrating the Lean Enterprise. Fall 2005. Lecture 10, "Lean Production Systems," defines Kanban as a signal that authorizes production or movement, establishing it as a generic signaling mechanism within a pull system. The application described in the question is a direct extension of this principle to a non-production process.

Multiple Linear Regression (MLR) is a statistical technique used to model the linear relationship between one dependent (output) variable, Y, and two or more independent (input) variables, X's. A primary application within methodologies like Six Sigma is to analyze historical data to understand which inputs significantly affect an output. If a strong predictive model can be built, it may reduce or eliminate the need for a more resource-intensive Designed Experiment (DOE). For the model to be valid and reliable, several assumptions must be met, including the critical assumption that the independent variables (X's) are not highly correlated with each other (i.e., they are independent, a condition known as no multicollinearity).

A. MLR is based on the assumption of linear relationships. Non-linear relationships require data transformation or different modeling techniques.
B. Quantifying uncertainty in the slope is a feature of all regression models (including simple linear regression), not a specific reason to use MLR.

1. For C & E: Montgomery

D. C.

Peck

E. A.

& Vining

G. G. (2012). Introduction to Linear Regression Analysis (5th ed.). Wiley.

Page 67

Section 3.1: "In multiple linear regression

we are interested in building a model that relates a dependent variable y to two or more independent or regressor variables

say x1

x2

...

xk." (Supports C)

Page 93

Section 3.3.3: Discusses the problem of multicollinearity

stating

"The situation where the regressors are highly correlated is called multicollinearity... If there is a near-linear dependency between the regressors

the least-squares estimates of the individual regression coefficients may be unstable." This underscores the assumption that X's should be independent. (Supports E)

2. For D: Antony

J.

& Banuelas

R. (2002). Key ingredients for the effective implementation of Six Sigma program. Measuring Business Excellence

6(4)

20-27. https://doi.org/10.1108/13683040210451679

Page 24

"Analyze Phase": The authors discuss the use of regression analysis to "identify the key process input variables ('the vital few x's') that affect the process output ('the y')." This analysis of existing data is presented as a precursor to the Improve phase

where a DOE might be used. A successful regression model can often provide enough insight to make improvements without a full DOE. (Supports D)

3. For C & E: The Pennsylvania State University. (n.d.). STAT 501: Regression Methods. Penn State Eberly College of Science.

Lesson 4.1: "Multiple linear regression is an extension of simple linear regression that allows for more than one independent variable." (Supports C)

Lesson 12.1: "Multicollinearity—or collinearity—is a situation in which two or more predictor variables are linearly related... The primary concern is that as the degree of multicollinearity increases

the regression model estimates of the coefficients become unstable." This reinforces the assumption of independent predictors. (Supports E)

In Lean methodology, the "Value Stream" refers to the complete set of all actions, both value-added and non-value-added, required to bring a product or service from its initial concept or raw material stage to the customer. This end-to-end flow includes the movement of materials and information. The primary objective of Lean is to analyze this entire sequence, typically through Value Stream Mapping, to identify and eliminate waste (muda), thereby improving efficiency and delivering maximum value to the customer. The question's description of "the sequence of activities and work required to produce a product or a service" is the precise definition of a value stream.

A. Value-add Flow: This is not a standard Lean term. The value stream explicitly includes both value-added and non-value-added activities to provide a complete picture for waste elimination.
B. Production Map: This is a generic term. The specific Lean tool used to visualize the value stream is a "Value Stream Map," but the flow itself is the "Value Stream."
D. Operating Procedure: This refers to a documented set of instructions for performing a single, specific task, not the entire end-to-end sequence of activities for a product.

1. Rother, M., & Shook, J. (2003). Learning to see: Value stream mapping to create value and eliminate muda. The Lean Enterprise Institute. On page 2, the value stream is defined as "...all the actions (both value added and non-value added) currently required to bring a product through the main flows essential to every product...".

2. Womack, J. P., & Jones, D. T. (2003). Lean thinking: Banish waste and create wealth in your corporation (2nd ed.). Simon & Schuster. In Chapter 1, "Value," the authors introduce the concept, stating, "The second key principle of lean thinking is to identify the value stream for each product..." and define it as the entire set of activities involved.

3. MIT OpenCourseWare. (2004). 16.852J / ESD.62J Integrating the Lean Enterprise, Fall 2005. Massachusetts Institute of Technology. Lecture 4 notes, "Value Stream Mapping," define the value stream as "The set of all actions (value-added and non-value-added) required to bring a specific product or service from raw material to the customer." (p. 4).

In the context of Lean Six Sigma and inferential statistics, analysis requires evaluating two distinct types of significance. Statistical significance, typically determined by a p-value, indicates whether an observed effect in the data is unlikely to have occurred by random chance. However, a statistically significant result may not be practical significance. Practical significance refers to the magnitude and real-world importance of the effect. A small, trivial improvement could be statistically significant but offer no meaningful business value. Therefore, Six Sigma practitioners must consider both to ensure that data-driven decisions are not only statistically sound but also operationally and financially beneficial.

A. Problematic: This is a descriptor of a situation's quality, not a formal type of significance used in statistical analysis.
B. Impractical: This is the antonym of practical significance and does not represent the statistical counterpart in hypothesis testing.
C. Usable: This term is too general. Both statistical and practical significance are criteria that help make data analysis "usable" for decision-making.

1. Munro

R. A.

Ramu

G.

& Zrymiak

D. J. (2015). The Certified Six Sigma Green Belt Handbook (2nd ed.). ASQ Quality Press. In the section on hypothesis testing

it is stated

"The project team must consider both statistical and practical significance when making a decision." (p. 259).

2. Ho

E. S. (2013). Statistical and practical significance: A case study of a six sigma project in a manufacturing environment. International Journal of Lean Six Sigma

4(4)

419-431. The abstract and body of this paper explicitly contrast the two concepts: "This paper aims to illustrate the difference between statistical significance and practical significance in a Six Sigma project..." https://doi.org/10.1108/IJLSS-09-2012-0003

3. Pennsylvania State University. (n.d.). STAT 500: Applied Statistics

Lesson 4: Inferences for the Mean of a Population. Eberly College of Science. The courseware distinguishes between the two

noting

"Statistical significance doesn't necessarily mean that the result is practically significant. The difference may be statistically significant but the difference may not be large enough to be of any practical importance."

The statement is a correct and complete definition of nominal scale data. Nominal data, also known as categorical or attribute data, involves classifying items into distinct categories identified by names or labels. These categories have no intrinsic numerical value or order. For example, classifying defects as "scratch," "dent," or "paint flaw" uses a nominal scale. Because there is no mathematical ranking, you cannot perform arithmetic operations like calculating an average or standard deviation. The only valid statistical measure for central tendency is the mode, which represents the most frequently occurring category.

B. False: This is incorrect. The statement accurately describes the fundamental characteristics of nominal data: it is categorical, lacks a mathematical order, and is unsuitable for arithmetic calculations beyond frequency counts.

1. Munro

R. A.

Ramu

G.

& Zrymiak

D. J. (2015). The Certified Six Sigma Green Belt Handbook (2nd ed.). ASQ Quality Press. In Chapter 11

"Measure: Basic Statistics

" the section on "Types of Data" states

"Nominal data (also called categorical or attribute data) arise when observations are classified into one of several mutually exclusive categories... Because the data are labels

any calculations are meaningless" (p. 138).

2. Pennsylvania State University. (n.d.). STAT 200: Elementary Statistics

Lesson 1: Introduction to Statistics. Eberly College of Science. In the section "Variables and Measures

" the courseware defines a nominal variable as one that "characterizes (or describes

or names) an element of a population... Arithmetic operations don't make any sense for nominal data."

3. Stevens

S. S. (1946). On the theory of scales of measurement. Science

103(2684)

677–680. This foundational paper defines the nominal scale as the most primitive form of measurement. It states

"The nominal scale represents the most unrestricted assignment of numerals... The only permissible mathematical operation is counting the number of cases in each class" (p. 678). https://doi.org/10.1126/science.103.2684.677

The statement is true. Capability analysis for attribute data (e.g., pass/fail, go/no-go) estimates the proportion of nonconforming items. Unlike continuous variable data, each attribute data point provides less information. To achieve a statistically sound and reliable estimate of the true process defect rate, a significantly larger sample size is required. This is particularly critical for processes with low defect rates (high capability), as a small sample may not capture any defects, leading to a misleading and overly optimistic capability assessment. A large sample ensures that the calculated proportion of defects is a stable and accurate representation of the process's long-term performance.

B. False: This is incorrect. Small sample sizes with attribute data lead to wide confidence intervals for the defect proportion, making the capability estimate highly uncertain and statistically unreliable for decision-making.

1. Montgomery

D. C. (2013). Introduction to statistical quality control (7th ed.). John Wiley & Sons.

Reference: In Chapter 8

Section 8.4

"Process Capability for Attribute Data

" the text explains that the estimate of the fraction nonconforming

p̂

is used. The statistical precision of this estimate is directly dependent on the total number of items inspected (n). A larger n is required to obtain a narrow confidence interval and thus a more reliable estimate of the true process fraction nonconforming.

2. The ASQ Certified Six Sigma Green Belt Handbook (2nd ed.). (2015). ASQ Quality Press.

Reference: Chapter 21

"Process Capability

" discusses the requirements for attribute capability studies. It emphasizes that for a meaningful analysis

the sample size must be large enough to ensure several nonconforming units are likely to be found. For example

if the expected defect rate is 1%

a sample of 50 is inadequate as it would likely yield zero defects.

3. Purdue University

Department of Statistics. (n.d.). STAT 51300: Statistical Quality Control Course Materials.

Reference: Lecture notes on "Process Capability for Attribute Data" typically state that the sample size (n) should be large enough such that np > 5 and n(1-p) > 5 for the normal approximation to the binomial distribution to be valid

which is a common method for calculating capability. This inherently requires a large sample size

especially when the proportion of defects (p) is small.

The 'Sustain' (Shitsuke) phase is the most challenging and is primarily driven by management. It focuses on instilling the discipline required to maintain the standards set by the first four S's, thereby integrating them into the organizational culture. This requires consistent leadership, setting expectations, conducting regular audits, providing resources for continuous improvement, and establishing recognition systems. While management must support all 5S activities, their direct, long-term commitment is the critical factor that ensures the gains are sustained and the system does not revert to its previous state.

A. Straighten: This is a hands-on organizational task primarily performed by the employees who use the specific workspace to arrange necessary items logically.
B. Sort: This is an initial clearing activity largely executed by the area's workforce to identify and remove unneeded items from the workplace.
C. Shine: This involves routine cleaning and inspection, which is a daily responsibility of the operators to maintain the cleanliness and order of their work area.

1. Gapp

R.

Fisher

R.

& Kobayashi

K. (2008). Implementing 5S within a Japanese context: an integrated management system. Management Decision

46(4)

565-579. On page 571

the authors state

"Shitsuke (sustain) is about making 5S a habit... This requires considerable management time and effort in coaching people and demonstrating the 5S principles on an ongoing basis." DOI: https://doi.org/10.1108/00251740810865067

2. Ho

S. K. (1999). Total quality management: an integrated approach. Kogan Page Publishers. In discussions of 5S as a foundational TQM practice

the text emphasizes that Shitsuke (Sustain) requires management to establish a formal system of audits and checks to prevent backsliding

making it a key management responsibility for cultural change (Section on 5-S Practice).

3. Liker

J. K. (2004). The Toyota Way: 14 Management Principles from the World's Greatest Manufacturer. McGraw-Hill. In Chapter 11

"Use Visual Control So No Problems Are Hidden

" the concept of 5S is explained. The text implicitly and explicitly links the sustainability of such systems to management's role in creating a culture of continuous improvement and discipline

which is the essence of the 'Sustain' phase.

4. Purdue University. (n.d.). Lean Manufacturing Tools: 5S. Purdue Manufacturing Extension Partnership (MEP). Courseware and training materials emphasize that the Sustain phase is about discipline and habit

which must be reinforced by management through performance reviews

audits

and consistent communication.

The statement provides a correct and fundamental definition of a population in the context of statistics. A population encompasses the entire collection of items, individuals, or, as stated, the data points associated with them that are of interest for a particular study or analysis. It represents the complete set from which a statistical sample is drawn. For example, if a study's interest is the diameter of all bolts produced in a day, the population is the collection of diameter measurements for every single bolt produced that day.

B. False: This option is incorrect. The provided definition is a standard and widely accepted description of a statistical population, representing the totality of observations or measurements for a group being studied.

1. Kubiak, T. M., & Benbow, D. W. (2009). The Certified Six Sigma Green Belt Handbook (2nd ed.). ASQ Quality Press. In Chapter 11, "Basic Statistical Terms," a population is defined as "a complete set of data" (p. 131).

2. Montgomery, D. C. (2013). Introduction to Statistical Quality Control (7th ed.). John Wiley & Sons. In Chapter 2, "The DMAIC Process," a population is described as "the entire collection of objects or outcomes about which information is sought" (p. 28).

3. Pennsylvania State University. (n.d.). STAT 200: Elementary Statistics. Eberly College of Science. In Lesson 1, "Introduction to Statistics," the courseware defines a population as "the entire group that you want to draw conclusions about." This group is represented by its data points.

The statement is correct. In a simple linear regression model, the equation is expressed as Y = β₀ + β₁X + ε. The term β₀ (beta-naught) is the constant, also known as the Y-intercept. It represents the predicted or mean value of the dependent variable (Y) when the independent variable (X) is equal to zero. Geometrically, it is the point where the regression line crosses the vertical (Y) axis. This term is fundamental for establishing the baseline of the model's prediction.

B. False: This is incorrect. The definition provided in the question is the standard and universally accepted definition of the intercept term (β₀) in a linear regression model.

1. Montgomery, D. C. (2013). Introduction to Statistical Quality Control (7th ed.). John Wiley & Sons. In Chapter 11, Section 11-2, "Simple Linear Regression," the model is defined as y = β₀ + β₁x + ε, where "β₀ is the intercept." (p. 476).

2. The Pennsylvania State University. (n.d.). STAT 501: Regression Methods, Lesson 2: Simple Linear Regression. Eberly College of Science. Retrieved from https://online.stat.psu.edu/stat501/lesson/2. The course material states, "β₀ is the population y-intercept of the regression line. It is the value of the mean of the response when the predictor is 0."

3. Siviy, J., & DeCarlo, N. (2005). An Introduction to Regression Analysis. Carnegie Mellon University, Software Engineering Institute. Technical Report CMU/SEI-2005-TN-033. In Section 2.1, "The Simple Linear Regression Model," the equation Y = β₀ + β₁X + ε is presented, and β₀ is explicitly defined as "the intercept of the line." (p. 4).

The Lean methodology is directly derived from the principles of the Toyota Production System (TPS), which was developed in Japan by Toyota Motor Corporation after World War II. Key architects like Taiichi Ohno and Shigeo Shingo established TPS to eliminate waste (muda), improve flow, and add value for the customer. Core concepts such as Just-in-Time (JIT), Jidoka (autonomation), and Kaizen (continuous improvement) are foundational to both TPS and Lean. The term "Lean" was coined by John Krafcik in 1988 and popularized by the book "The Machine That Changed the World" to describe this highly efficient production philosophy.

A. Honda: While Honda has its own efficient production system, it is not the origin of the Lean philosophy.
C. Ford: The Ford system of mass production was an important precursor, but Lean was developed specifically to overcome its rigidity and waste.
D. Motorola: Motorola is credited with developing the Six Sigma methodology in the 1980s, which focuses on reducing variation and defects.

1. Womack

J. P.

Jones

D. T.

& Roos

D. (1990). The Machine That Changed the World. Rawson Associates. In Chapter 4

"The Rise of the Lean Production System

" the authors explicitly identify the Toyota Production System as the foundation of Lean.

2. ASQ. (2022). The Certified Six Sigma Green Belt Handbook

Third Edition. ASQ Quality Press. Chapter 4

"Introduction to Lean

" states

"Lean has its roots in the Toyota Production System (TPS)."

3. Krafcik

J. F. (1988). Triumph of the Lean Production System. Sloan Management Review

30(1)

41–52. This is the first peer-reviewed publication to use the term "Lean" to describe the system pioneered by Toyota.

4. Ohno

T. (1988). Toyota Production System: Beyond Large-Scale Production. Productivity Press. This book

written by a primary architect of TPS

serves as a foundational text explaining the system from which Lean principles were derived.

Regression analysis is a statistical method used to model the relationship between a dependent variable (output, Y) and one or more independent variables (inputs, X). Its application is justified under several conditions. A core assumption of simple linear regression is that the relationship between variables is linear (A). A primary purpose of developing a regression equation is to understand how inputs influence the output, which is fundamental for process control (C). The resulting model can also be used for prediction, which is especially valuable when direct measurement of the output is costly, destructive, or time-consuming (E). Within a problem-solving framework like DMAIC, regression is a key analytical tool used to identify the significant root causes of defects and guide improvement efforts (D).

B. If a relationship is known to be non-linear, applying a standard linear regression model is inappropriate and will result in a poor fit and invalid conclusions.

1. Montgomery

D. C. (2013). Introduction to statistical quality control (7th ed.). John Wiley & Sons.

Page 468

Section 11-2.1: States the simple linear regression model is Y = β₀ + β₁x + ε

which explicitly defines a linear relationship. (Supports A).

Page 467

Section 11-2: "Regression models are a very useful tool for process control and improvement... If we can build a model that relates the process output variable to the important controllable input variables

then we can use this model to determine the values of the input variables that will optimize the process." (Supports C and D).

Page 474

Section 11-2.2: Discusses using the fitted model ŷ = β̂₀ + β̂₁x for prediction. (Supports E).

2. Munro

R. A.

Ramu

G.

& Zrymiak

D. J. (2015). The Certified Six Sigma Green Belt Handbook (2nd ed.). ASQ Quality Press.

Page 267

Section "Regression": "Regression analysis is a statistical technique for modeling the relationship between variables... The goal is to be able to predict the value of a dependent variable based on the value of at least one independent variable." (Supports C and E).

Page 269

Section "Assumptions in Linear Regression": Lists "Linearity" as the first assumption

stating "The relationship between X and the mean of Y is linear." (Supports A).

The handbook presents regression as a tool within the DMAIC methodology

which is fundamentally a structure for solving problems such as process defects. (Supports D).

3. Pennsylvania State University. (n.d.). STAT 501: Regression Methods

Lesson 1: Simple Linear Regression. Penn State Eberly College of Science. Retrieved from https://online.stat.psu.edu/stat501/lesson/1

Section 1.1 - What is Simple Linear Regression?: "In general

a regression model allows us to... describe the linear relationship between the predictor x and the response y... [and] predict the value of a response for a given value of a predictor." (Supports A

C

and E).

When data from two or more distinct populations are combined into a single data set, the resulting frequency distribution will often exhibit multiple peaks, or modes. Each peak typically corresponds to the central tendency of one of the underlying populations. A bimodal distribution, characterized by two distinct peaks, is the classic and most common example of this phenomenon. It serves as a strong indicator that the data is not homogeneous and may be stratified, originating from different sources such as different machines, shifts, operators, or materials that were assumed to be the same but are performing differently.

A. Skewed: Skewness describes asymmetry in a distribution. While mixing populations can sometimes result in skew, a skewed distribution can easily arise from a single, unmixed source (e.g., income data).
C. Gaussian: A Gaussian (or normal) distribution is symmetric with a single peak. It represents data from a single, stable process, which is the opposite of the scenario described.
D. Tri-peaked: This is a specific case of a multimodal distribution with three peaks. While possible with three sources, "bimodal" is the more general and fundamental term used to illustrate this concept.

1. ASQ Quality Press. T.M. Kubiak & Donald W. Benbow, The Certified Six Sigma Green Belt Handbook, 2nd ed., 2015. In Chapter 15, "Basic Statistical Tools," the section on Histograms states, "A bimodal distribution...often indicates that the data are from two different processes" (p. 239).

2. Pennsylvania State University. Eberly College of Science, STAT 500: Applied Statistics Course Notes. In Lesson 1.3.3.1 - Shape, it is noted, "A bimodal distribution has two peaks. A bimodal distribution is often a mixture of two different unimodal distributions (i.e., a mixture of two different populations)."

3. University of Virginia Library. Research Data Services + Sciences, "Getting Started with Histograms." The guide explains, "A histogram is bimodal if it has two peaks. This often means the data is from two different populations."

Statistical Process Control (SPC) charts are a fundamental tool for monitoring process stability and are, by definition, time-series plots. The horizontal axis (x-axis) represents the progression of time or the sequence in which samples are taken (e.g., sample number, date, or lot number). By standard graphical convention, this sequence progresses from left to right. Therefore, data points are plotted chronologically, with the oldest data appearing on the left and the most recent data point being added to the far right-hand side of the chart. This layout is essential for identifying trends, shifts, and patterns as they emerge in real-time.

B. False: This is incorrect. The primary purpose of an SPC chart is to monitor a process over time. Reversing the time axis would make the chart counterintuitive and difficult to interpret for timely process control.

1. Montgomery

D. C. (2020). Introduction to Statistical Quality Control (8th ed.). Wiley.

Page 190

Section 5.2.1: "A control chart is a time-sequenced chart... The vertical axis represents the sample statistic

and the horizontal axis represents the sample number or time." The accompanying figures throughout the chapter (e.g.

Figure 5-1) consistently show the sample number increasing from left to right.

2. NIST/SEMATECH. (2012). e-Handbook of Statistical Methods. National Institute of Standards and Technology.

Section 6.3.1

"What is a Control Chart?": The handbook states

"The control chart is a graph used to study how a process changes over time. Data are plotted in time order." All graphical examples provided

such as the one in Section 6.3.2

Figure 6.3

show the "Sample Index" on the x-axis increasing from left to right.

3. Oakland

J. S. (2008). Statistical Process Control (6th ed.). Butterworth-Heinemann.

Page 113

Chapter 6: "The results of sampling and testing are plotted on a graph with the variable on the vertical axis and the sample number or time on the horizontal axis... This simple but powerful tool is the control chart." This description confirms the standard left-to-right time progression.

An interaction occurs when the effect of one input variable (factor) on an output (response) is dependent on the level of another input variable. In this scenario, the combined effect of the two inputs is different from the sum of their individual effects. It is possible for the main effects of the individual inputs to be statistically insignificant, while their interaction effect is highly significant. This is a fundamental concept in Design of Experiments (DOE) and analysis of variance (ANOVA).

B. Oddity: This is a general, non-technical term and does not have a specific statistical meaning relevant to this context.
C. Coincidence: This term implies a random or chance occurrence, whereas an interaction is a systematic and repeatable effect that can be statistically modeled.
D. Impossibility: This phenomenon is not an impossibility; it is a well-documented and common occurrence in complex systems studied through statistical methods like DOE.

1. Montgomery

D. C. (2017). Design and Analysis of Experiments (9th ed.). John Wiley & Sons.

Section 5-3

The Two-Factor Factorial Design with a Fixed Effects Model

p. 174: "In many experiments

we find that the difference in the response between the levels of one factor is not the same at all levels of the other factors. When this occurs

there is an interaction between the factors."

2. NIST/SEMATECH. (n.d.). e-Handbook of Statistical Methods.

Section 5.3.3.5

Interaction Effects: "An interaction effect exists if the effect of one variable on the response depends on the level of the other variable(s)... It is possible for the main effects to be small

but the interaction effect to be large." Retrieved from https://www.itl.nist.gov/div898/handbook/pri/section3/pri335.htm

3. The ASQ CSSGB Certification Preparation

ASQ Six Sigma Green Belt Council.

Measure Phase

Design of Experiments (DOE): The official ASQ Body of Knowledge for the CSSGB exam includes "main effects and interaction plots" as a key topic. The concept is defined as the phenomenon where factors working in combination produce results that are different from the sum of their separate effects.

4. Pennsylvania State University. (n.d.). STAT 502 - Analysis of Variance and Design of Experiments. Eberly College of Science.

Lesson 5: Factorial Designs

Section 5.1 - The 2-Factor Factorial Design: "An interaction is the failure of one factor to produce the same effect at different levels of another factor." The course material provides graphical examples of strong interactions with weak main effects.

This situation describes a process that is not capable of meeting customer requirements. Control limits are derived from process data (the "voice of the process") and represent its natural variation. Specification limits are determined by the customer (the "voice of the customer"). When the natural variation of the process (control limits) is wider than the customer's acceptable range (specification limits), the process will produce non-conforming output even when it is stable and in a state of statistical control. This guarantees defects and is a highly undesirable state, indicating a low process capability index (Cp < 1.0).

A. The lower Control Limit for the R chart is equal to zero This is a common and expected occurrence for range charts with small subgroup sizes (n ≤ 6), as the control chart constant D3 is zero. It is not an undesirable situation.
B. Attempt to use SPC for tracking transaction times at a warehouse SPC is a versatile tool applicable to both manufacturing and transactional processes. Tracking metrics like transaction times is a valid and beneficial use of SPC to monitor process stability and performance.
C. A process is in Statistical Control before implementation of SPC This is an ideal, not undesirable, situation. Implementing SPC on a stable process allows for the establishment of a valid performance baseline, monitoring for future deviations, and focusing on process improvement.

1. Montgomery

D. C. (2012). Introduction to Statistical Quality Control (7th ed.). John Wiley & Sons.

For Correct Answer (D): In Chapter 8

Section 8-2.1

"Process Capability Analysis Using a Histogram or a Probability Plot

" Montgomery explains the relationship between natural tolerance limits (approximated by control limits) and specification limits. On p. 367

it is stated

"If the natural tolerance limits are wider than the specification limits... the process is not capable of meeting specifications." This directly supports that this is an undesirable situation.

For Incorrect Option (A): In Appendix VI

Table VI

"Factors for Constructing Variables Control Charts

" the constant D3 used to calculate the LCL for the R chart is listed as 0 for subgroup sizes n=2 through n=6. This confirms that an LCL of zero is a standard calculation result.

2. The ASQ Certified Six Sigma Green Belt Handbook (2nd ed.). (2015). ASQ Quality Press.

For Correct Answer (D): Chapter 23

"Process Capability

" discusses the comparison of the voice of the process to the voice of the customer. It explicitly states that when the process spread is greater than the specification spread

the process is not capable

which is an undesirable outcome requiring improvement.

For Incorrect Option (B): Chapter 20

"Statistical Process Control (SPC)

" provides examples and discusses the application of SPC beyond manufacturing

including service and transactional environments

confirming its suitability for such applications.

3. Purdue University

School of Industrial Engineering. (n.d.). IE 336: Production and Operations Control - Lecture Notes on Statistical Process Control.

For Incorrect Option (C): Course materials on SPC implementation typically state that the first step is to bring a process into statistical control. Starting with a process already in control is presented as the desired state from which to monitor and begin continuous improvement activities

not as a problem.

In Lean methodology, "Conveyance," more commonly known as Transportation, is one of the eight forms of waste (muda). It refers to any unnecessary movement of materials, products, or information from one location to another. This movement does not add value for the customer but consumes resources such as time, labor, and energy, and increases the risk of damage to the goods. The primary goal in a Lean system is to minimize this waste by optimizing the physical layout of the workspace, creating a more streamlined process flow, and locating processes closer to one another.

A. Overprocessing: This waste involves adding more value to a product than the customer requires, such as unnecessary features or overly tight tolerances.
B. Motion: This waste refers to the unnecessary movement of people or equipment within a process, not the movement of materials between processes.
D. Correction: This is the waste associated with reworking defective products or fixing errors, which consumes resources to correct a problem.

1. Earley, J., & An, J. (2012). Lecture 2: Lean Fundamentals. MIT OpenCourseWare, 16.660J Introduction to Lean Six Sigma Methods. Slide 17, "The 8 Wastes," defines Transportation as "Moving products that are not actually required to perform the processing." [Available at: https://ocw.mit.edu/courses/16-660j-introduction-to-lean-six-sigma-methods-iap-2012/pages/lecture-notes/]

2. The Ohio State University, Fisher College of Business. (n.d.). The 8 Wastes of Lean. Center for Operational Excellence. This resource defines Transportation waste as the "Unnecessary movement of materials, products, or information."

3. Hines, P., & Rich, N. (1997). The seven value stream mapping tools. International Journal of Operations & Production Management, 17(1), 46-64. https://doi.org/10.1108/01443579710157989. On page 52, the article identifies "transportation" as a key non-value-adding activity, describing it as the movement of materials from one point to another.

Fractional factorial designs are a type of screening experiment. They are most effective in the initial stages of process investigation when there are many potential factors to consider, but little is known about their individual effects or how they interact. These designs efficiently identify the most significant factors (the "vital few") with a minimal number of experimental runs. This efficiency is achieved by making a calculated assumption—based on the principle of sparsity of effects—that higher-order interactions (e.g., three-way or higher) are negligible. This approach is ideal for exploring a system where the key relationships are not yet understood.

A. Much is known: If much is known about the interactions, a more detailed experiment like a full factorial or a response surface design would be more appropriate to precisely model those known relationships.
C. We don’t care: This is inaccurate. We make a strategic assumption that higher-order interactions are negligible to gain efficiency, but we often still care about and can analyze main effects and some two-way interactions.
D. Data exists: The existence of historical data does not dictate the choice of an experimental design. DOE is used to generate new, structured data to establish cause-and-effect, which existing data may not provide.

1. Montgomery, D. C. (2017). Design and Analysis of Experiments (9th ed.). John Wiley & Sons. In Chapter 8, "Two-Level Fractional Factorial Designs," the introduction states, "When the number of factors is large, the number of runs required for a full or complete factorial design... can become very large... In this situation, it is often useful to use a fractional factorial design... These are screening experiments." (p. 275).

2. Antony, J. (2014). Design of Experiments for Engineers and Scientists (2nd ed.). Elsevier. Chapter 5, "Fractional Factorial Experiments," explains, "Fractional factorial designs are good for screening... The primary objective of such experiments is to select or screen out the few important main effects from the many less important ones." (p. 61).

3. The Pennsylvania State University. (n.d.). STAT 503: Design of Experiments. In Lesson 10: Two-Level Fractional Factorial Designs. Retrieved from https://online.stat.psu.edu/stat503/lesson/10. The course material notes, "Fractional factorial designs are a good choice for screening experiments. These are experiments in which many factors are considered with the purpose of identifying those factors that have large effects." This context implies a situation where knowledge is limited and requires screening.