Module 5 - Six Sigma Quality Analysis & Process Improvement Flashcards
What is process outcome
How the output of the process is characterized
A “stable” process is one that is: __ and __
unchanged over time and in statistical control
Note: There will still be random variation
A capable process is one whose outcomes:
meet its goals or objectives
Analyzing process capability is impossible if it is __, because…..
the process is unstable because such processes are unpredictable
Why do decision makers in services often react unnecessarily to random variation and what harm does this cause? (explain)
Humans are not naturally good at making decision in the presence of random variation. Humans see changes in outcomes and immediately assume that the process has changed. When reacting to random variation, we may initiate a search for root cause of a change when, in fact, no change has occurred. We may end up changing a perfectly good process. This action could disrupt employees’ ability to serve customers and eventually spread “fear” within an organization.
How is acceptance sampling performed? (describe)
Name a con.
(1) A sample of parts is inspected to determine how many conform to design specifications.
(2) Based on these results, the batch of parts from which the sample was drawn was either passed on to the next stage of production or the entire batch was subjected to a “100% inspection.”
Cons: not designed to find the root cause of poor quality
Total Quality Management (TQM) is __ oriented
It was developed by __
process
Demming
Why is acceptance sampling a poor process improvement tool?
Because it is done after the process is run (often much later), it is typically done by inspectors who may not be familiar with the process, and it does not maintain the order of production. Therefore, it is ineffective at identifying root cause of problems.
DPMO
defects per million opportunities
Why would the Six Sigma quality metric be ineffective as a universal quality goal?
Because not all products are similar in complexity to those found at Motorola. Some applications (e.g., high volume with a high cost of defects) may warrant a stricter goal, while other applications (e.g., simpler commodity products) may warrant a less strict goal.
MBO
management by objectives
Statistically speaking, why is it difficult to create a fair bonus system based on process outcomes?
Because of the impact of random variation of outcomes. An outcome may not meet the objective even though the process is acceptable (and vice-versa).
What’s the formula for customer satisfaction?
Satisfaction = performance - expectations
Process outcomes can be __ or __
Tangible or intangible
Two important concepts in the statistical analysis of quality are __ & __
process stability
&
process capability
According to Deming, when the MBO approach is used in a quality system, it is flawed for a number of reasons.
1-it does not consider the impact of random variation.
2-it usually leads to managers gaming the process
3-it fails to recognize that performance is mostly dictated by the system, not the individual employees.
What’s the difference between a stable and capable process?
Stable - process is predictable (unchanging and statistically controlled)
Capable - process meets its goals
Statistical control charts were developed by: __
The basis of these rules is: __
It is __ oriented
They are used to determine if a process is __
They recognize that outcomes are subject to __
Walter Shewhart
normal distribution
process-oriented
stable
random variation
___ was based on the process characteristics found at Motorola
It was developed to replace ___
six sigma
TQM (Total Quality Management)
Today, six sigma seeks to make improvements based on projects whose benefits are __ __
Financially quantifiable
For control charts,
UCL =
Center Line =
LCL =
UCL = µ + 3σ
Center Line = µ = p-bar
LCL = µ - 3σ
σ formula on control chart
what is p?
what is n?
what is UCL and LCL?
Note: a proportion could never be less than __
σ = sq.rt { [ p * (1 - p) ] / n }
where
p = the likelihood of success
n = the sample/subgroup size
UCL = p + 3σ
LCL = p - 3σ
0%
The most popular Shewhart rules for detecting an out of control process are:
- 1 point exceeding any 3-sigma limit
- 2 of 3 consecutive points beyond 2-sigma limits
- 4 of 5 consecutive points beyond 1-sigma limits
- 8 consecutive points on the same side of the center line
- 10 of 11 consecutive points on the same side of the center line.
On a control chart, would it be a good idea to show the objective as a horizontal line?
No. The presence of such a line would imply that the organization’s goal is to have each outcome exceed the objective, which is a de-facto MBO approach.
On a P Chart, if the sample size were increased, the control limits: __
If the sample size decreased, the control limits: __
Why?
would become more narrow
would become wider
If you take larger samples from each batch (e.g., 50M items), you get a more accurate picture of the true proportion of defective items in that batch. The control limits on the chart would be narrower because you have a more precise estimate
Can a process be in control but making a significant number of nonconforming parts?
Yes. When a process is in control (i.e., stable) it may still be off target and/or it may have a high level of natural variation relative to the product specifications.
Six Sigma addressed a problem stemming from increased __
How does OFD relate?
product complexity
When a product becomes more complex, the number of OFDs increases. (OFD = opportunities for defects)
Six Sigma seeks to maximize profits by __ __ to eliminate defects.
It focuses on __ increases as criteria for success
reducing variation
revenue
Six Sigma uses a DMAIC structure. This stands for:
D - define the problem (and everything else about the project)
M - measure the space (gather info, describe current process)
A - analyze the current process to find where improvements should be made
I - improvement ideas (a list of them to solve the problem)
C - control (action plan to confirm & sustain improvements)
Name the following for Lean and Six Sigma for:
1 - type and length of project
2 - project focus is to improve __ by __
3 - project is selected based on which will:
4 - analysis techniques are geared towards
5 - begins with:
Lean:
1- quick Kaizen projects in a few days; team works full time
2- customer service by minimizing wasteful practices
3- removes wastes, add value for customers
4- intuitive & descriptive displays, root cause analysis, mistake proofing
5- brainstorming session
Six Sigma:
1- deliberate projects over months; formal/disciplined DMAIC structure
2- customer service by minimizing variation & errors
3- decrease costs or increase revenue
4- statistical data analysis, controlled experiments, optimization
5- executive education & extensive training
In general:
- Lean is simple, quick and removes waste or adds value. Qualitative.
- Six Sigma is long, complex, and removes costs or adds revenue. Quantitative.
Each Lean Six Sigma project should include three stages
Project Initiation, Project Execution, and Project Communication
These generally mean what?
µ
σ
mean
standard deviation
Control charts are used to determine if:
If so, it will show a __ pattern
a process is stable
random
If a process does not achieve six sigma quality,
its variation must be reduced or its specifications must be relaxed
True or False: If a company makes a product with 6-sigma quality, then these products would have a defect rate no worse than 3.4 defects per million products sold?
False. The 6-sigma metric applies to each OPPORTUNITY for defect, NOT the final PRODUCT
Lean Six Sigma combination team should be a group of __ members
6-12
What is DMAIC used for?
It’s a framework for managing a project
