Question 1

Statistical Control of Product Quality 
2 possible approaches:

Accepted Answer

1.statistical process control (SPC)
2.acceptance sampling (inspection)

Question 2

SPC is preferable to acceptance sampling where possible:

Accepted Answer

1.Prevention, rather than detection 2.
More timely feedback 3.Inspection is a non-value-adding activity, but it costs a firm time and money

Question 3

acceptance sampling

Accepted Answer

applies statistical sampling techniques to the decision whether to 1.“accept” a production lot (and pass it on to the customer) or 2.to “reject” a production lot (and either scrap the lot or sort the lot and rework the defective parts)

Question 4

random sampling

Accepted Answer

every unit in the population (production lot) has an equal chance of being included in the sample

Question 5

any time we rely on sample information to make a decision about an entire lot, there is

Accepted Answer

some margin for error

Question 6

two possible types of errors in acceptance sampling

Accepted Answer

1.accept a “bad” lot (consumer’s risk,Beta symbol) 2. Reject a “good” lot (producer’s risk,alpha symbol)

Question 7

Statistical Process Control
Basic idea:

Accepted Answer

control process parameters and the product will take care of itself

Question 8

Statistical Process Control
Primary tools:

Accepted Answer

1. process control charts 2.developed by Shewhart 3.signal the operator when “special cause” variation is present (i.e., the process is out of control)

Question 9

sources of “special cause” variation:

Accepted Answer

1.equipment – tool wear, machine vibrations, electrical fluctuations 2.	material – thickness, moisture content 3. 
environment – temperature, light, humidity
4.operator – physical and emotional well-being, lack of understanding

Question 10

Statistical Process Control
Types of control charts:

Accepted Answer

1.variables data – a precise measurement of some physical dimension is taken 2.attributes data – the number or proportion of units with some characteristic is tracked

Question 11

variables data – kinds of charts

Accepted Answer

1. X and R charts 2. X and s charts

Question 12

attributes data – kinds of charts

Accepted Answer

1. p chart 2. c chart

Question 13

Steps 1-4  Constructing X and R Charts

Accepted Answer

1.Choose sampling plan 2.Calculate the range (R) and the mean (X) for each sample 3.Calculate the average range (R) and the grand mean (X) for all samples 4.Calculate the average range (R) and the grand mean (X) for all samples

Question 14

Step 1 - considerations for choosing the  sampling plan

Accepted Answer

1.at least 25-30 samples (more is better) 2.sample typically includes 4-5 units of product 3.critical for process to be “in control” when the data is collected for charts 4.Outliers for which the cause is known are thrown out

Question 15

constructing X and R Charts - use the Control Chart Factors Table

Accepted Answer

table will lead to control limits approximately three standard deviations from the centerline values

Question 16

See slides 11-16 for Constructing X and R Charts: An Example

Accepted Answer

example

Question 17

See slides 17-20 for Control Chart Factors Table, Control Limit Calculations. Set up Control Charts and Track process as production occurs

Accepted Answer

example

Question 18

statistical process control - sp.c.v. means special cause variation

Accepted Answer

1.control charts monitor output of process so we know if sp.c.var. is present 2.sp.c. var. suspected when non-randomness in data 3.if sp.c.var. present, identify & eliminate the cause of  variation-prevent recurrence of the problem-process improvement

Question 19

non-randomness in the data:

Accepted Answer

1.“runs” of values steadily above or below the center line 2.upward or downward trend
3.periodicity 4.any points outside the control limits

Question 20

tolerance limits

Accepted Answer

tell us whether a particular item is acceptable or defective (e.g., 3.000”  0.002”)

Question 21

control limits

Accepted Answer

control limits – allow us to track process mean and variation over time; no individual item measurements are recorded

Question 22

tolerance limits and control limits

Accepted Answer

1. are not the same 2.so, a process could be in control but have low capability (i.e., producing too many defects) or could be out of control and producing no defects

Question 23

Controlling Process Variation - 5 things to consider

Accepted Answer

1.some level of variation in all processes 2.How much process variation is too much? 3.How do we measure amount of process variation? 4.determine if process is highly capable 5. higher process capability, consistently high quality products

Question 24

Controlling Process Variation
some level of variation is inherent in all processes – really, no such thing as two “identical” products 
How much process variation is too much?

Accepted Answer

1.it depends on how much the critical dimensions of the product are allowed to vary without affecting product performance – a design issue 2.EX.: product specifications or tolerance limits for ruler length 12.000” ± 0.002”

Question 25

How do we measure the amount of process variation?

Accepted Answer

for a stable process, we can measure the standard deviation of the critical dimensions of the output (product) 




A higher process capability will lead to more consistently high quality products

Question 26

high process capability



A higher process capability will lead to more consistently high quality products

Accepted Answer

if the amount of variation inherent in the process is small relative to the allowable variation in the product

Question 27

formula for process capability

Accepted Answer

process capability equals acceptable variation divided by actual variation

Question 28

some level of variation in all processes

Accepted Answer

no two identical products

Question 29

Process Capability Example (See slide 25) and HANDOUT from 11.06

Accepted Answer

check it out

OTM 17

Quality Part 2

Question	Answer
Statistical Control of Product Quality 2 possible approaches:	1.statistical process control (SPC) 2.acceptance sampling (inspection)
SPC is preferable to acceptance sampling where possible:	1.Prevention, rather than detection 2. More timely feedback 3.Inspection is a non-value-adding activity, but it costs a firm time and money
acceptance sampling	applies statistical sampling techniques to the decision whether to 1.“accept” a production lot (and pass it on to the customer) or 2.to “reject” a production lot (and either scrap the lot or sort the lot and rework the defective parts)
random sampling	every unit in the population (production lot) has an equal chance of being included in the sample
any time we rely on sample information to make a decision about an entire lot, there is	some margin for error
two possible types of errors in acceptance sampling	1.accept a “bad” lot (consumer’s risk,Beta symbol) 2. Reject a “good” lot (producer’s risk,alpha symbol)
Statistical Process Control Basic idea:	control process parameters and the product will take care of itself
Statistical Process Control Primary tools:	1. process control charts 2.developed by Shewhart 3.signal the operator when “special cause” variation is present (i.e., the process is out of control)
sources of “special cause” variation:	1.equipment – tool wear, machine vibrations, electrical fluctuations 2. material – thickness, moisture content 3. environment – temperature, light, humidity 4.operator – physical and emotional well-being, lack of understanding
Statistical Process Control Types of control charts:	1.variables data – a precise measurement of some physical dimension is taken 2.attributes data – the number or proportion of units with some characteristic is tracked
variables data – kinds of charts	1. X and R charts 2. X and s charts
attributes data – kinds of charts	1. p chart 2. c chart
Steps 1-4 Constructing X and R Charts	1.Choose sampling plan 2.Calculate the range (R) and the mean (X) for each sample 3.Calculate the average range (R) and the grand mean (X) for all samples 4.Calculate the average range (R) and the grand mean (X) for all samples
Step 1 - considerations for choosing the sampling plan	1.at least 25-30 samples (more is better) 2.sample typically includes 4-5 units of product 3.critical for process to be “in control” when the data is collected for charts 4.Outliers for which the cause is known are thrown out
constructing X and R Charts - use the Control Chart Factors Table	table will lead to control limits approximately three standard deviations from the centerline values
See slides 11-16 for Constructing X and R Charts: An Example	example
See slides 17-20 for Control Chart Factors Table, Control Limit Calculations. Set up Control Charts and Track process as production occurs	example
statistical process control - sp.c.v. means special cause variation	1.control charts monitor output of process so we know if sp.c.var. is present 2.sp.c. var. suspected when non-randomness in data 3.if sp.c.var. present, identify & eliminate the cause of variation-prevent recurrence of the problem-process improvement
non-randomness in the data:	1.“runs” of values steadily above or below the center line 2.upward or downward trend 3.periodicity 4.any points outside the control limits
tolerance limits	tell us whether a particular item is acceptable or defective (e.g., 3.000”  0.002”)
control limits	control limits – allow us to track process mean and variation over time; no individual item measurements are recorded
tolerance limits and control limits	1. are not the same 2.so, a process could be in control but have low capability (i.e., producing too many defects) or could be out of control and producing no defects
Controlling Process Variation - 5 things to consider	1.some level of variation in all processes 2.How much process variation is too much? 3.How do we measure amount of process variation? 4.determine if process is highly capable 5. higher process capability, consistently high quality products
Controlling Process Variation some level of variation is inherent in all processes – really, no such thing as two “identical” products How much process variation is too much?	1.it depends on how much the critical dimensions of the product are allowed to vary without affecting product performance – a design issue 2.EX.: product specifications or tolerance limits for ruler length 12.000” ± 0.002”
How do we measure the amount of process variation?	for a stable process, we can measure the standard deviation of the critical dimensions of the output (product) A higher process capability will lead to more consistently high quality products
high process capability A higher process capability will lead to more consistently high quality products	if the amount of variation inherent in the process is small relative to the allowable variation in the product
formula for process capability	process capability equals acceptable variation divided by actual variation
some level of variation in all processes	no two identical products
Process Capability Example (See slide 25) and HANDOUT from 11.06	check it out

"Know" box contains:
Time elapsed:
Retries: