The DMAIC Lean Six Sigma Project and Team

The DMAIC Lean Six Sigma Project and Team Tools Approach Measure Phase 1

Lean Six Sigma Combo/Black Belt Training! Agenda – Measure Phase Welcome Back, Brief Review Process Thinking, Mapping, and Analysis Measurement System Analysis Sigma Level, Baseline Metrics, Types of Data Capability Analysis Introduction to Minitab Pareto Analysis Theories of Xs and Cause and Effect Data Collection Plan and Sampling Lessons Learned / Measure Phase Conclusions Wrap-Up / Teach-Coach Practice / Quiz 2

Measure Objectives (pg. 8 -11) • Identify the Project Y • Define the performance standards for Y, and its baseline (current state) performance • Clarify understanding of specification limits as well as defect and opportunity definitions • Validate the measurement system (MSA) • Collect the data as needed • Characterize the data using basic tools and capability • Begin funneling the X’s that affect the Y • Measure…what is the current state/performance level and potential causes 3

Why spend so much time in the Measure phase? “When you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meager and unsatisfactory kind…” Lord Kelvin “If you can’t measure it, you can’t manage it. ” Peter Drucker 4

Why Do We Measure? • To thoroughly understand the current state of our process and collect reliable data on process inputs that you will use to expose the underlying causes of problems • To know “where you are” – the extent of the problem • To understand quantify the critical inputs (xs) that we believe (theories) are contributing to our problem (Ys) 5

Lean Six Sigma DMAIC Phase Objectives • Define… what needs to be improved and why • Measure…what is the current state/performance level and potential causes • Analyze…collect data and test to determine significant contributing causes • Improve…identify and implement improvements for the significant causes • Control…hold the gains of the improved process and monitor 6

LSS PROJECT FOCUS Process Characterization Define The right project(s), the right team(s) Y Analyze Process Optimization Process Problems and Symptoms Process outputs Response variable, Y r Measure r r r q q Improve Control q X’s q q Independent variables, Xi Process inputs The Vital Few determinants Causes Mathematical relationship Goal: Y = f ( x ) 7

Measure Phase: Process Mapping 8

The Basic Philosophy of Lean Six Sigma • • All processes have variation and waste All variation and waste has causes Typically only a few causes are significant To the degree that those causes can be understood they can be controlled • Designs must be robust to the effects of the remaining process variation • This is true for products, processes, information transfer, transactions, everything • Uncontrolled variation and waste is the enemy 9

Remember - What is Six Sigma… • A high performance measure of excellence • A metric for quality • A business philosophy to improve customer satisfaction • Focuses on processes and customers • Delivers results that matter for all key stakeholders • A tool for eliminating process variation • Structured methodology to reduce defects • Enables cultural change, it is transformational 10

Why Process Thinking? Allows criticism without blaming people Allows shared understanding of how things work Helps manage complexity Provides focus within context Helps to manage scope of project Identification of team members Understand inputs / outputs - leads to measurement 11

High Level Process Map - SIPOC Process Name Supplier-Inputs-Process-Outputs-Customer …. ……………………………………………. . . ………………………………………………… 12

High Level 1 Box Examples Inputs Customer Name Customer ID Bill to Ship to Credit status Quoting Job Outputs Time to quote Number of contacts Quote accuracy 13

High Level Process Flow INPUTS PROCESS OUTPUTS Specialty available Chart available Patient assessment MD orders consult Order in chart—complete Reason for consult Order flagged Order placed in correct area Legible order Computer system working Unit Sec enters consult Consult stamp on chart Consult documented in CERNER Contact information Call schedules Assigned vs. Group call schedule Unit Sec calls consult Specific MD notified Answering service notified MD on-call notified 24 hour chart check RN reviews chart for completeness Consult not met Failure to meet consult is noted by RN 24 hr chart check signature RN realizes need to reconsult RN informs Unit Sec to reconsult Unit Sec attempts to reconsult Contact information Call schedules Assigned vs. Group call schedule Unit Sec/RN verifies with exchange / office Office or exchange notifies physician 14

Lean Six Sigma Project and Team Basic Tools Process Flow Chart (pg. 33 -44) A visual display of the key steps and flow of a process, also called a process map. Usually standard symbols are used to construct process flow charts. These include boxes (or rectangles) for specific steps, diamonds for decision points, ovals for defined starting and stopping points, and arrows to indicate flow. Processes can include providing a service, making or delivering products, information sharing, design, etc. – Should represent the current as-is state of the process! 15

Process Mapping (pg. 33 -44) • A process is a sequence of steps or activities using inputs to produce an output (accomplish a given task). • A process map is a visual tool that documents and illustrates a process. • Several styles and varying levels of detail are used in Process Mapping. Most common and useful styles are SIPOC, Flow Diagrams, Box Step, and Value Stream Maps. 16

Process Mapping • The team should start with the observed, current, as-is process. • Start high-level, and work to the level of detail necessary for your project (key inputs). • As inconsistencies are discovered, the team can develop a future state or shouldbe process map to improve the key xs and the overall output (Y) of the process. 17

Levels of Process Mapping How Low Can You Go? Level 1: Core Business Processes Level 2: Processes Level 3: Subprocesses Level 4: Activities/Steps Level 5: Task 18

Patient Care Core Business Process Admissions Treatment & Invervention Discharge Medication administration Physical therapy Diagnostic and therapeutic imaging intervention Lab testing Cardiology treatment intervention Pulmonary treatment intervention Surgical intervention IV therapy treatment Nutritional support Discharge teaching Physiological monitoring Implementation of treatments Communication Pain management Billing 19

How Low Can Should You Go? • Decompose the process until it becomes unnecessary to go any farther – Accountability is identified – Responsibility falls outside the process boundaries – Root cause becomes evident – The time required to measure the process exceeds the time required to perform it 20

Flow Diagrams - Concept (For Complete List, see: Power. Point - Shapes - Flowchart) Activity / Step Connector Decision Off-page Connector Flow lines Database Terminal / End Document 21

Process Flow - Symbols Follow the standard symbols; don’t make up your own. People who follow your process flow should be able to understand your work and documents. 22

Don’t mess with it. Hide it! YES Does the thing work? NO Does anyone know? NO YES You poor dummy! NO Can you blame someone else? Did you mess with it? NO YES You big dummy! Will you get in trouble? NO YES NO PROBLEM Toss it! 23

Suspected Bleeding Disorder Sample Process Map H & P yes Definitive Family History? Focused Testing yes (see list a) no no Symptomatic Patient or Family History? END yes Screening Tests: CBC PT PTT PFA Thrombin Time Positive Test Result? no yes Further Testing Required? no Confirmed Dx yes Positive Test Result? Release or worku for other Dx no Positive yes Focused Testing Screening Test (see list b) Result? Review Screening no Test Results 24

http: //www. qualitym ag. com 25

Process Flow Chart Lean Six Sigma Project Selection A Gap Exists Define Potential Project Draft Problem Statement Redefine Project Scope Identify the Metrics No Reconsider Project Determine the Outputs (Y) Two Or Fewer Outputs? No Yes Charter and Launch Project Yes Meets Six Sigma Criteria? Calculate Benefits Quantify the Opportunity

http: //www. oregon. gov 27

A Flow Chart of Process Mapping Start Assemble the Team Define the Process Scope Macro Map? Yes Create Process Flow Diagram Identify VA/NVA Steps Find the Hidden Factory Revise and Update Observe and Verify List Process Capability Build a Detailed Map Identify the Specs. Identify X’s and Y’s No Draft a Macro Map Tools: Power. Point, Excel, Visio, Process Model 28

Additional Process Mapping Techniques • Swim lanes (pgs. 43 -44) • Value stream mapping (pgs. 45 -51) • Time Value Map (pgs. 52 -53) 29

Process Mapping Analysis Detailed Analysis of Process Delays or Errors: Identifying process delays or potential errors is an important analyze phase activity. Going into greater detail in identifying the type and source of delay or error will help to more clearly define the root cause and thereby produce a more robust solution and overall improvement. 30

Process Mapping Analysis Types of Process Delays or Errors: • • 31 Gaps Redundancies Implicit or unclear requirements Bottlenecks Hand-offs Conflicting objectives Common problem areas

Process Mapping Analysis Gaps – Responsibilities for certain process steps are unclear, not understood, easy to “skip” – Process seems “unfocused, ” goes off track in delivering what the customer needs – Excessive variation 32

Process Mapping Analysis Redundancies – Actions or steps are duplicated – Different groups repeat actions that are done somewhere else, and they are not aware of the repeat actions occurring – Excessive checking (non-value adding) 33

Process Mapping Analysis Implicit or unclear requirements – “Word of mouth” instructions, not formally documented; assumptions – Operational definitions are noted; different groups interpret definitions and instructions differently – Unclear measurement system 34

Process Mapping Analysis Bottlenecks – A “slow down” of work flow – Multiple inputs may feed into a process step, which is then delayed – Output of entire process may be “controlled” by the output rate of the bottleneck step(s) 35

Process Mapping Analysis Hand-offs – Unclear if a process step has received needed inputs from an “upstream” step – Misunderstanding of who is responsible, or who has done what – Communication problems 36

Process Mapping Analysis Conflicting objectives – Unclear alignment from one group to another working in the same process – Direction from leadership and metrics – Communication problems 37

Process Mapping Analysis Common problem areas – Overall weaknesses seen throughout a process, common failure modes – Repeated steps or checks in a variety of places throughout the process flow – Communication problems – The “Hidden Factory” 38

The Hidden Factory All of the work that is performed that is above and beyond what is required to deliver good products and services to the customer; work that is not necessarily tracked (cost, productivity, etc. ). Work-arounds or “built-in” Rework 39

Process Mapping, Measurement and Analysis Study your key processes and note any of the aforementioned potential process delays or errors directly on your process map. Go to the source to verify with data. Many key xs are identified through careful and deliberate process measurement and analysis. 40

Process Map Analysis Frequency of VS checks? Start Can Cerner flag critical VS changes? Ongoing assessment and monitoring of patients vital signs and status Patient medical record Cerner data Change in patient’s physical status Handoff issues? NO Continued deterioration YES NO Potentially bad clinical outcome Use of MRTs? NO Kaizen bursts identify handoffs or transactions that have the potential to create defects 41 Did we recognize change? Nursing skill to recognize shock? Did we act quickly? YES Does a full ICU mean delays? Was the action appropriate? YES Appropriate care delivered Best possible outcome Are we effectively communicating vital info?

42

Measure Phase: Measurement System Analysis (MSA) Can the variation in the parts (output) be detected over and above the variation caused by the measurement system? 43

Baseline Data Questions • What is the current process capability? (Where are we now in terms of consistently meeting the customer’s needs? ) • Is the process stable? • How much improvement do you need to meet your goal, to make a meaningful impact? • What data are currently available? • How will you know whethere has been an improvement? • How does the current state compare to the CTQs? 44

Measurement System Analysis (MSA) (pgs. 87 – 103) Is it the right data to answer the question at hand? or Is it the best question the existing data can answer? 45

Look Carefully 46

Measurement System Analysis (MSA) (pg. 87 – 103) A measurement system analysis is performed to determine if the measurement system can generate true reliable data, and to assure the variation observed is due to the actual performance of the process being studied, and not due to excessive variation in the measurement system itself. 47

Measurement System Analysis (MSA) “In any program of control we must start with observed data; yet data may be either good, bad, or indifferent. Of what value is theory of control if the observed data going into that theory are bad? This is the question raised again and again by the practical man (woman). ” - Walter Shewhart 48

Reliable Data ? 49

Separate what we think is happening from what is really happening! 50

Data Integrity? • What assumptions were made? • Is the data representative of the process ? • Who generated the data? • How was it measured? • What is the noise in the measurement? • If required, does it pass an audit? • Can we trust the data and the measurement system used to generate the data to properly investigate the process? 51

Inspection Exercise: You have 60 seconds to document the number of times the 6 th letter of the alphabet appears in the following text: The Necessity of Training Farm Hands for First Class Farms in the Fatherly Handling of Farm Live Stock is Foremost in the Eyes of Farm Owners. Since the Forefathers of the Farm Owners Trained the Farm Hands for First Class Farms in the Fatherly Handling of Farm Live Stock, the Farm Owners Feel they should carry on with the Family Tradition of Training Farm Hands of First Class Farmers in the Fatherly Handling of Farm Live Stock Because they Believe it is the Basis of Good Fundamental Farm Management. 52

6 Items To Look For In A Good Measurement System üResolution üConsistency üRepeatability üReproducibility üLinearity üAccuracy 53

Resolution • Is the measuring base unit small enough to adequately evaluate the variation in the process? • Can we “see” differences in what the process is producing? • Must monitor the process frequently enough to catch it varying, or going from good to bad. • As a general rule, we should use units of measure that are at least 10 subdivisions of the range of measurement being investigated. “Ten bucket rule” Examples of issues with resolution in your projects? 54

Consistency (Stability) Issue • Does the measurement system error remain stable or predictable over time, across equipment, across operators, across all shifts, across all facilities, etc…? • Will we get reliable measurements from the process even if the measurements are taken on the weekends, during night shifts, by different employees, etc. ? 55

Measurement Systems Would it be OK if the time clock your employees get paid by is off by: 1 hour every day? 1 hour a week? 1 hour per month? 1 hour per year? Measurement Systems must be Repeatable & Reproducible if we are to draw adequate conclusions 56

Repeatability / Precision • The variation in measurements obtained when one operator uses the same measuring process for measuring the identical characteristic of the same parts or items ( part dimension, blood pressure cuff, chemistry analyzer, etc. ). • Can the variation in the parts be detected over and above the variation caused by the measurement system? • How closely will successive measurements of the same part or process by the same person using the same instrument repeat themselves? 57

Reproducibility • The variation in the average of measurements made by different operators using the same measuring process when measuring identical characteristics of the same items (two abstractors reviewing same chart). • Reproducibility is very similar to repeatability. The primary difference is that instead of looking at the consistency of one person, we are looking at the consistency between people. • Are the average measurements for each part reproducible across different operators, gages, machines, locations, etc…? 58

Linearity • Is the measurement system consistent across the entire range of the measurement scale? • Are measurements reliable even at the extremes? 59

Accuracy • Are the measurements truly representative of the output of the process being studied? • On average, do I get the “true data” from the output of the process? 60

Accuracy vs. Precision. . . . Not Accurate, Not Precise. . Accurate but not precise . . . . Precise but not accurate Accurate and Precise 61

62

Key Questions for a MSA? (Your Project’s Measurement System) • Is my measurement system repeatable - will I get the same results if I take the measurement more than once? • Is my measurement system reproducible - will someone else be able to complete the same measurement and get the same results? • Is my measurement system accurate - will the results from my study match the actual value, or expert data? 63

MSA Recap INADEQUATE Most of the variation is accounted for by physical or actual differences in the process or components. - All sources of measurement variation will be small - You can have higher confidence that actions you take in response to the data are based on reality Variation in how the measurements are taken is high. - You can’t tell if differences between units or process observations are due to the way they were measured, or are true differences - You can’t trust your data and therefore shouldn’t react to perceived patterns, special causes, etc. —they may be false signals 64

Why do we conduct MSA? (Your Project’s Measurement System) • While many statistical tools may be very powerful, they can also provide misleading results if there is too much measurement error. • We conduct MSA to gain an understanding of the quality, or trustworthiness, of data being collected to drive decisions about improving your process(es). • Some part of the total observed variation inherent to a process is, in fact, caused by the measurement system itself. – How much variation can we tolerate? • A good measurement system is vital for your baseline data as well as your investigations of possible Xs. 65

Measure Phase: Calculating Sigma Levels and Baseline Data and Metrics 66

Why are Baseline Measures so Important? “If we could first learn where we are and where we are going, we would be better able to judge what to do and how to do it. ” Abraham Lincoln 67

Calculating the Approximate Sigma Level 1. Define your opportunities 2. Define your defects 3. Measure your opportunities and defects 4. Calculate your yield 5. Look up process Sigma 68

Calculating the Approximate Sigma Level Define your opportunities and defects • An opportunity is any area within a product, process, service, or other system where a defect could be produced or where you fail to achieve the ideal product or service in the eyes of the customer . • A defect is any type of undesired result. The defect threshold may be as superficial as whether or not the product works. But it may be more subtle. – This may be the difference between “Does the car run? ” and “Does the car have a flawless paintjob, the tires I want, the brand of CD changer I want, etc…” – It’s usually not enough just to ask whether the product “meets expectations”… the expectations need to be defined. 69

Calculating the Approximate Sigma Level Measure your opportunities and defects and calculate your yield – the percent without defects. Opportunities - Defects Opportunities x 100 Total number of widgets minus widgets with defects Total number of widgets 156 183 x 100 85. 24% 70

Calculating the Approximate Sigma Level • Look up process Sigma A 85. 24% yield is a process Sigma of 2. 5 to 2. 6 Discussion: What is your estimate of your process Sigma 71

Activity Working individually 1. Define an opportunity in your process. What’s a ballpark estimate of the number of opportunities in your process? 2. Define the defects in your process. What’s a ballpark estimate of the number of defects in your process? 3. Calculate your process yield Opportunities - Defects Opportunities x 100 4. Find your Sigma level (10 minutes to complete) 72

Balancing Measures • Balancing measures are often identified to prevent important process, input, or output factors from being sacrificed at the expense of achieving a narrow goal. • Prevent “tunnel-vision” • Be alert for unintended consequences • “Need to know” versus “nice to know” • Balancing measures are those things we don’t want to lose sight of as we drive toward meeting our goal. 73

Introductory Statapult Activity! Working in teams, • – – – Try to hit a target distance (specification) with a projectile of your choice and your assigned statapult Collect the distance for each shot by team member in sequential order (6 total shots for each team member) In addition to the actual distance shot, also record if the shot is “in spec”, or “out of spec” Collect and record the total time it takes each team member to complete their respective 6 shots List potential xs that explain variation in the distance the projectile travels (Y) (If you have any variation? ) List any waste that occurred in your statapult process How well did your team perform? What is your team’s sigma level? Are you individually a good statapultician? 74

Baseline Data Questions • What is the current process capability? • Is the process stable? • How much improvement do you need to meet your goal? • What data are currently available? How many samples do I need to collect (pg. 85 -86) • How will you know whethere has been an improvement? • How does the current state compare to the CTQs? 75

Types of Data Two major types of data (pg 70) – Continuous (or “variable”) • Measurement along a continuum, length, height, age, time, dollars, etc. – Discrete (or “attribute”) • Categories, yes/no, names, labels, counts, etc. 76

Types of Data Continuous – Any variable that can be measured on a continuum or scale that can be infinitely divided – There are more powerful statistical tools for interpreting data continuous data, so it is generally preferred over discrete/attribute data – Examples: height, weight, age, respiration rate, etc. 77

Types of Data Discrete Data Type Definition Example Count How many? Count of errors; How many patients got evidencebased care? How many specimens were tested? Binary Data that can have only one of Was delivery on-time? Was the two values product defect-free? Alive/dead; Male/female; Yes/No Nominal The data are names or labels with no intrinsic order or relative quantitative value Colors; dog breeds; diagnoses; brands of products; nursing units; facility Ordinal The names or labels represent some value inherent in the object or item (there is an obvious order to the items) Product performance: excellent, very good, fair, poor; Severity: mild, moderate, severe, critical 78

Types of Data Example: Type of data: • Product meets design specifications • Discrete – Binary • Heart rate • Continuous • Distribution managers • Discrete – Nominal • Gasoline grades (regular, plus, premium) • Discrete - Ordinal 79

Baseline Capability • A baseline capability study basically answers how well the current “as is” process meets the needs (specifications) of the customer. It can be tracked over time via run chart, control chart, etc. • Process Capability compares the output of a process to the needs of the customer for a given key measure. 80

Process Capability Uncontrolled Variation is Evil Traditional Philosophy Taguchi Philosophy “goalpost mentality” LSL USL Anything outside the specification limits represents quality losses LSL USL Any deviation from the target causes losses to the business 81

Process Capability: Variation The New Goalpost Scoring The New Business Reality 3 Points 2 Points 1 Point 82

Characteristic of the Performance Gap… (Problem) Accuracy and/or Precision Off-Target Variation LSL USL On-Target Center Process Reduce Spread LSL USL LSL = Lower spec limit USL = Upper spec limit The statistical approach to problem solving 83

Process Capability: Short Term and Long Term Short Term Long Term | | | -5 -4 -3 -2 -1 0 1 2 3 4 5 84

Process Capability: Short Term and Long Term • Processes experience more variation over a longer term than in the short term. • Capability can vary depending on whether you are collecting data over a short term or a long term. • The equations and basic concepts for calculating capability are identical for short term and long term except for how standard deviation is calculated to account for the increased variation over the long term. 85

Is a 3 s process a capable process? Long-term Capability LSL Perfect World – Accurate & Consistent USL Consistent, but not always accurate Time Short-term Capability 86

Process Capability: Short Term and Long Term • Short Term (Cp and Cpk calculations) – Gathered over a limited number of cycles or intervals – Gathered over a limited number of shifts & associates • Long Term (Pp and Ppk calculations) – Gathered over many cycles, intervals, equipment, & operators – May be attribute or variable – Assumes the data has “seen” at least 80% of the total variation the process will experience 87

Process Capability: Short Term and Long Term (pgs. 135 – 140) • Cp (short term) and Pp (long term) calculations compare the amount of variation in the process output to the total range of variation allowed (customer specifications) 88

A Problem With Cp and Pp Which is the better process? What is the difference in Cp between the two processes? | | | -5 -4 -3 -2 -1 0 1 2 3 4 5 What can be done to make Cp more effective as a process capability statistic? | | | -4 -3 -2 -1 0 1 2 3 4 5 89

Process Capability: Short Term and Long Term (pgs. 135 – 140) • Cpk (short term) and Ppk (long term) compares the amount of variation and the location of the mean from the process output to the total range of variation allowed (customer specifications) 90

Meet Ppk / Cpk Process Performance Example: A process mean is 355, standard deviation is 15, upper spec. limit is 380, and lower spec. limit is 270 What is the Cpk? What is the Cp? | | | -4 -3 -2 -1 0 1 2 3 4 5 91

Capability – Cpk’s Centered Process Cpk = USL-Mean Cp = USL – LSL 3 s 6 s OR Cpk = Mean – LSL USL m Shifted Process Cp = same C = pk pk less LSL m 3 s less USL LSL m USL 92

Cpk and Process Sigma USL LSL -6 -5 -4 -3 LSL -2 -1 0 +1 +2 +3 +4 USL LSL Cpk = 1. 67 +/- 3σ within spec limits +/- 5σ within spec limits +5 +6 USL -6 -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 +6 LSL USL Cpk = 1. 33 Cpk = 2 +/- 4σ within spec limits +/- 6σ within spec limits 93 -6 -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 +6 -6 -5 -4 -3 -2 -1 0+1 +2+3+4+5+6

Run Charts The Importance of Data Over Time Continuous Y (e. g. Length of Stay) Graphical display: Run charts (also called. Time-series charts) average Discrete X (e. g. Month) 94

Data Analysis / Statistical Software: Minitab Brief Overview 95

Improving how we Improve! (Through Data Analysis and Minitab) Minitab is a tool consisting of many tools and techniques for thorough data analysis. 1. Do not think of Minitab as “giving you the answer. ” 2. If you do not have reliable data, and/or you are not asking the proper analysis questions, Minitab will be of little value – if any! 96

Improve: Data-Driven Approach Is there a difference between Data and Information? Data – factual information used as a basis for reasoning Information – the communication or reception of knowledge obtained from investigation, study, or instruction 97

Minitab • Typical desktop icon for Minitab 98

Minitab Overview Toolbar Session Window Test results and messages will appear as running text. The text in this window can be modified, copied, and pasted Worksheet You can have multiple worksheets with your data arranged in columns. The grey line is where you put your column labels 99

Minitab Overview 100

Text column 101 Date column Numeric data column

Data Analysis and Minitab Remember the triple C’s for Data in Minitab 1. Organize data into Columns 2. Record/Input data Chronologically as appropriate 3. Data must be Clean (no commas, dollar signs, etc. ) 102

Descriptive Statistics • Using the data collected in the statapult exercise, look at the descriptive stats – Stat>Basic Statistics>Display Descriptive Statistic – Stat>Basic Statistics>Graphical Summary 103

Descriptive Stats Descriptive Statistics: Distance Variable Distance N 75 Variable Distance Maximum 87. 000 104 N* 0 Mean 78. 880 SE Mean 0. 549 St. Dev 4. 756 Minimum 55. 000 Q 1 77. 000 Median 79. 000 Q 3 81. 000

Graphical Summary 105

Capability Analysis • Stat>Quality Tools>Capability Analysis 106

Short Term Variation - Example • Use Minitab to estimate short term variation: – Stat > Quality Tools > Capability Analysis (Normal)

Capability Six-Pack 108

Measure Phase: Pareto Charting and Analysis (The 80/20 Rule) 109

Pareto chart • A Pareto chart is a special type of bar graph where the categories are arranged from largest to smallest with a line indicating the cumulative percent Vilfredo Pareto observed that 80% of the land in Italy was owned by 20% of the population. Later, Joseph Juran called this “ 80 -20 rule” the Pareto principle. 80% of the effects come from 20% of the causes. 110

Lean Six Sigma Project and Team Basic Tools Pareto Analysis (pg. 142 -144) A Pareto chart is simply a bar graph with the bars arranged typically in descending order from highest to lowest frequency by discrete category. It graphically displays the 80/20 rule. Approximately 80% of the quantifiable results (frequency), will be attributed to 20% of the causal categories. 111

Create the Pareto Chart • • Go to Stat>Quality Tools>Pareto Chart Select “Chart Defects Table” Defects or attribute data in: Colors Frequencies in: Counts 112

Create the Pareto Chart • Click on Options • Label the X axis “M&M Color” • Label the Y axis “Count” • Give your chart a title • Click on OK again 113

Your Pareto Chart …should look something like this: 114

Lean Six Sigma Project and Team Basic Tools 115

Measure Phase: Cause and Effect Analysis (Collecting the “theories” of x’s) 116

Statapult Activity Follow-up • Working with your team – – – – 117 Discuss the effect (Y results) of your statapult process (the head of your fishbone diagram)? How satisfied are you with the measurement system for your process output List some potential xs (theories) that affect your process outcome (Y). Construct a fishbone diagram of the potential x’s Discuss how we might determine the most significant x’s List some categories of waste experienced by your team Prepare a mini-presentation (5 mins) to share with class

Lean Six Sigma Project and Team Basic Tools Cause and Effect Diagrams (pg. 146 -149) A C&E diagram (also called a fishbone diagram), is a pictorial display of the potential or likely causes of a given effect. The causes are grouped and arranged in meaningful categories, sometimes called branches. There are numerous ways to name the grouped branches. The most common names include: Material, Method, Manpower, Machinery, Measurement, and Mother Nature (Environment). 118

119

120

Lean Six Sigma Project and Team Basic Tools 121

Other Fishbone categories • 6 Ms – Method, Material, Manpower, Machinery, Measurement, Mother Nature • 4 Ps – Policies, Procedures, Personnel, Place 122

Cause & Effect Matrix Form Natural break, Sanity check 123

Cause and Effect Chart • Stat>Quality Tools>Cause-and-Effect • In Minitab, you can build your C&E Chart from lists of potential Xs in the workbook or by keying them into the dialogue box 124

Xs in the Worksheet 125

Xs typed in as constants 126

Sub-branches 127

Measure Phase: Data Collection Plan and Preparation for Analysis (Data Collecting for the “theories” of x’s) 128

Data Collection Plan (pgs. 72 – 81) • Data are the documentation of an observation or measurement. Data are facts, but you may need information – data which provide the answers to questions you have. • A good data collection plan helps ensure data will be useful (measuring the right things) and statistically valid (measuring things right). 129

Data Collection Plan (pgs. 72 – 74) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 130 Decide what to collect Decide on stratification factors as needed Develop operational definitions Determine the appropriate/needed sample size Identify the source/location of data Develop data collection forms/check sheets Decide who will collect the data Train data collectors Do ground work for analysis Execute your data collection plan

Data Collection Plan 1. Formulate the question or theory: What is the question we are trying to answer? 2. Decide how data will be communicated analyzed. 3. Decide how to measure: population or sample? 4. Collect data with a minimum of bias. 131

Data Collection Plan Asking the Right/Best Question Time to ABX in Minutes is captured using a continuous measure: “How many minutes did it take? ” What kind of data will you be collecting? 132 It can be converted into a discrete measure: “Was it done within four hours? ”

Data Collection Asking the Right Question Is the measure you are using a good one? • Understandable • Provides information for decision making • Applies broadly • Is conducive to uniform interpretation • Is economical to apply • Is compatible with existing design of sensors • Is measurable even in the face of abstractions 133

Data Collection Plan Communicating the Results • Although you may not know what the data reveals – and it may seem odd to be thinking about how your team will analyze and display the data -- having some idea about the sort of analysis and display you will use will help you make decisions about the data you collect. • If you wait until after the data are collected to think about analysis, you may find that the data do not support the kind of analysis you want to conduct. 134

Sampling Qualities of a Good Sample • Free from bias – Bias is the presence of some undue influence on the sample selection process that causes the population to appear different than it actually is • Representative – The data should accurately reflect a population. Representative sampling helps avoid biases specific to segments of the population • Random – The data are collected in no predetermined order and each element has an equal chance of being selected 135

Sampling • Random Sampling – each element has an equal chance of being selected – Simple random (no pattern) – Systematic random (every Nth value) • Stratified Random Sampling – the population is grouped into levels or “strata” according to some characteristic and proportional samples are drawn randomly from each stratum 136

Random Sampling X X X X Sample X X Population 137 X X X Each element has an equal chance of being chosen

Stratified Random Sampling XXXXX YYYYY YY ZZZZZ Population 138 Sample • Randomly sampled from each stratified category or group • Sample sizes for each stratum are generally proportional to the size of the group within the population

Sampling The following are NOT appropriate ways to get a valid random sample: • Fixed percentage sampling – leads to undersampling from small populations and oversampling from large populations • Judgment sampling – using judgment to select x number of “representative” samples - guess • Chunk or convenience sampling – selecting sample simply because the items are conveniently grouped 139

Sampling (pgs. 85 -86) Sample size calculation for continuous data n = 1. 96 s 2 Δ 140 n Minimum sample size 1. 96 Constant representing a confidence interval of 95% (valid when sample size is 30 or more) s Estimate of standard deviation of data Δ The level of precision desired from the sample you are trying to detect (same units as s)

Sampling Sample size calculation for discrete data n = 1. 96 s Δ 141 2 P (1 -P ) n Minimum sample size 1. 96 Constant representing a confidence interval of 95% (valid when sample size is 30 or more) s Estimate of standard deviation of data P Estimate of the proportion defective Δ The level of precision desired from the sample you are trying to detect (same units as s)

Effective Data Driven Practice Steps to Effective Data Driven Practice Ask the Right Question - Bias the question with existing belief system Right/Appropriate Data - No easy access to data systems - Substitute what is needed with what is available - Missing and incomplete data - Data values are incorrect Proper Analysis Correct Audience Correct Interpretation Appropriate Action 142 Potential Failure Modes - Insufficient statistical skill - Inadequate statistical software - Analysis paralysis - Unable to take action - Decision errors from false positives / false negatives - Refusal to accept the facts - Bias the interpretation with existing belief system - Intellectual dishonesty - Unwilling to take action - Analysis paralysis

Lean Six Sigma DMAIC Phase Objectives • Define… what needs to be improved and why • Measure…what is the current state/performance level and potential causes • Analyze…collect data and test to determine significant contributing causes • Improve…identify and implement improvements for the significant causes • Control…hold the gains of the improved process and monitor 143

Project Name: Project Scope: Champion: Name Process Owner: Name Black Belt: Name Green Belts: Enter scope description Names Problem Statement: Mislabeled example Define Start Date: Enter Date End Date: Enter Date Measure Start Date: Enter Date End Date: Enter Date Analyze Start Date: Enter Date End Date: Enter Date ¨ Benchmark Analysis ¨ Identify Project Y(s) ¨ Identify Vital Few ¨ Project Charter ¨ Identify Possible Xs Root Causes of ¨ Formal Champion (possible cause and Variation Sources & Approval of Charter effect relationships) Improvement (signed) ¨ Develop & Execute Opportunities ¨ SIPOC - High Level Data Collection Plan ¨ Define Performance Process Map ¨ Measurement Objective(s) for Key q Customer CTQs System Analysis Xs ¨ Initial Team meeting ¨ Establish Baseline ¨ Quantify potential $ (kickoff) Performance Benefit ¨ Not Complete 144 ü Complete v Not Applicable Customer(s): CTQ(s): Defect(s): Beginning DPMO: Target DPMO: Estimated Benefits: Actual Benefits: Improve Control Start Date: Enter Date End Date: Enter Date ¨ ¨ ¨ Implement Sustainable Process Controls – Validate: § Control System § Monitoring Plan § Response Plan ¨ System Integration Plan q $ Benefits Validated q Formal Champion Approval and Report Out Generate Solutions Prioritize Solutions Assess Risks Test Solutions Cost Benefit Analysis ¨ Develop & Implement Execution Plan ¨ Formal Champion Approval Directions: • Replace All Of The Italicized, Black Text With Your Project’s Information • Change the blank box into a check mark by clicking on Format>Bullets and • Numbering and changing the bullet. Author: Enter Name Date: 30 September 2020

Going Forward with your Project and Analysis “What’s different in me is that I still pose to myself the questions that people quit making when they were five years old. ” Albert Einstein 145