Chapter 14
Consequences of too much variation
-Parts don't fit correctly -Products don't function properly -The company gets a reputation for poor quality
****Look at stable vs not stable homework problems
...
Statistical thinking is based on what 3 concepts?
1. A production system focuses on interconnectedness of processes and variables 2. Variation is everywhere 3. Firms need to understand variation to reduce it
Corrective action - what do you do when a process is "out of control?"
1. Carefully identify the quality problem 2. Form correct team to evaluate and solve problem 3. Use structured brainstorming w/ fishbone or affinity diagrams to identify causes of problems 4. Brainstorm to identify potential solutions 5. Eliminate the cause 6. Restart the process 7. Document the problem, root causes, and solutions 8. Communicate results of the process to all personnel so that process becomes reinforced and ingrained in the organization -if you have an out of control data point, you don't necessarily need to make a change *you first investigate and adjust if appropriate *if an investigation doesn't turn up anything unusual, measure another sample
Process control charts
1. Variable- continuous measurement (weight, height, volume, etc.) 2. Attribute- A defect is present or it's not -Process Control Charts can be based on variables or attributes
Not stable control charts
1. one data point out, above or below 2. two data points near upper or lower limit 3. five successive data points above/below central line 4. seven successive data points on increasing or decreasing line 5. erratic behavior 6. sudden change in level
Process stability
A process has stability if its variation is random and doesn't have a assignable cause -Process charts determine if a process has stability -So normal and non-normal variation help define process stability... -process control chart - visually can see what a process is doing and how it is behaving
Statistical thinking is based on all of the following concepts EXCEPT: A. statistics can explain every problem B. variation is everywhere C. firms need to understand variation to reduce it D. a system of production focuses on the interconnectedness of processes and variables
A. statistics can explain every problem
A world-class, or Six Sigma, capability index is A. 2.0. B. 1.0. C. 1.33. D. 1.25.
A. 2.0
Random variation
A.K.A. Common cause variation -Consistent mean and dispersion -Uncontrollable -Amount of variation can be large or small -Our statistical tools are NOT designed to detect random variation -common cause is normally occurring -hard to eliminate, takes design changes (product or process) -rarely negative
Nonrandom variation
A.K.A. Special cause variation -Inconsistent mean and dispersion -Caused by a specific event (change in work shift, etc) -It is special cause related and in all instances special -Causes should be understood and eliminated before even considering the common cause variation -This is part of the so called low hanging fruit in process and product improvements
Attributes control charts
Attributes deal with defects and defectives instead of continuous measurements -attributes are things like defects, good/bad, go/no-go Defects are countable and there could be many within the same unit- tracked -A defective unit, as a whole, does not meet performance requirements
Systematic samples are taken according to A. time or size. B. time or sequence. C. size or sequence. D. None of the above.
B. time or sequence.
In statistical process control, if a machine breaks due to poor maintenance, the process will likely display what kind of variation? A. random B. special cause C. abnormal D. common cause
B. special cause
Which of the following is NOT cause for investigation in a control chart? A. one data point above the upper control limit B. five successive data points above the central line C. two data points near the lower control limit (2 standard deviations beyond the mean) D. four successive points on a decreasing line
D. four successive points on a decreasing line
If a process has stability, its variation is ________ and does not have an _______ cause. A. random; unassignable B. non-random; unassignable C. non-random; assignable D. random; assignable
D. random; assignable
Which of the following is NOT a type of attributes control chart? A. p (proportion defective) B. np (number defective or number nonconforming) C. R (range) D. u (number defects per unit)
R (range)
Statistical process control charts
SPC Charts are used to monitor process variation -chart is created based on historical data, not based on customer specks or process specks
_______ are tough fits for control charts because...
Services because of customer-induced variation
Statistical process control
The application of statistical concepts to the production process to see if your processes display stability (they only exhibit random variation)
Applying control charts
There are many indicators of an out-of-control situation -Very likely our variation is not random ****in all instances the reaction is investigate, not adjust the process -look at pg. 438 (table 14.15) in the text
Process capability studies are used to determine whether a process is in need of monitoring through process control charts. A. True B. False
True
x(bar) and R Charts
We monitor measurements of a product to track 2 primary variables: mean and dispersion x(bar) charts track the average (mean) of the characteristic being measured R charts track the dispersion (range) of the process
Do control charts apply to all situations?
no
Should you adjust a process when only random variation is present?
no, this is tampering and by doing so, you introduce non-random variation
CpK
process capability index -standard index to state the capability of the process... -The higher the index the better, or more capable a process is in producing products that meet specifications... -Cpk use specifications limits in its calculation, and recall a control -Charts limits are calculated using process data and not specs -Because it uses spec limits it can be used to estimate yield (scrap)
Process capability
the ability of a process to produce a product that meets specification **once it's stable, make sure it's capable -so once special causes have been eliminated/controlled you should have a stable process and Calculating process capability is possible and useful.