Term
| What are the steps in measuring a process? |
|
Definition
1. Identify what to measure
2. Determine how to measure it
3. Develop a sampling plan and reaction plan
4. validate the measurement system
5. add to overall control plan |
|
|
Term
| What is used to identify what to measure |
|
Definition
| cause and effect matrix to identify which x input has the greatest y output impact |
|
|
Term
| how to determine how to measure the characteristic |
|
Definition
-establish and operational definition
-make data collection method
-train |
|
|
Term
| how to develop sampling and reaction plan |
|
Definition
-when and where to collect data and what to do with it
-what to do with out of spec conditions and points |
|
|
Term
| how to validate measurement system |
|
Definition
-gauge r&r to quantify error
-analyze stability, bias, linearity, repeatability, reproducibility |
|
|
Term
|
Definition
| Prioritizes process inputs (X) that have been identified as contributing the effect (Y) |
|
|
Term
| What is measurement system analysis (MSA) |
|
Definition
| Evaluates whether data is a representation of what is really happening |
|
|
Term
|
Definition
| Capacity of MSA to produce the same values over time |
|
|
Term
|
Definition
| distance between measurement and actual value |
|
|
Term
|
Definition
measure of consistency of bias.
a scale is off 1 pound for 150 lb person but 5 pounds off for 200 lb person |
|
|
Term
|
Definition
| same appraiser gets same values with same device |
|
|
Term
|
Definition
| variation between appraisers |
|
|
Term
| what does gauge r&r stand for? |
|
Definition
| repeatability and reproducibility |
|
|
Term
|
Definition
| getting different readings from the same appraiser-usually due to poor equipment or calibration |
|
|
Term
| Steps to complete gauge R&R study |
|
Definition
1. satisfy requirements
2. Prepare
3. collect samples
4. measure samples
5. analyze data
6. view results |
|
|
Term
| how to satisfy prerequisites? |
|
Definition
| Make sure stability, linearity, and bias is good |
|
|
Term
|
Definition
define sample size (p)
define number of trials (n)
define appraisers (o) |
|
|
Term
| what does it mean when p value is less than the specified alpha? |
|
Definition
| the variation is significant |
|
|
Term
| when do we want p value to be greater than alpha? |
|
Definition
|
|
Term
|
Definition
| measures the noise as it relates to the system |
|
|
Term
| what do we want to see in the x bar chart? |
|
Definition
| plot-points outside of the limits |
|
|
Term
| what does a range chart do? |
|
Definition
assesses 3 attributes of measurement system
-statistical control
-appraiser consistency
-resolution |
|
|
Term
| What does it mean if gauge R&R is >30%? |
|
Definition
| unacceptable, measurement process needs improvement |
|
|
Term
| What does it mean if gauge R&R is between 10 and 30% |
|
Definition
| marginal, system needs improvement |
|
|
Term
| What does it mean if gauge R&R is <10% |
|
Definition
| measurement system is acceptable |
|
|
Term
| What is the data collection error-measurement? |
|
Definition
| instrument errors, includes calibration errors |
|
|
Term
| What is the data collection error-operational? |
|
Definition
|
|
Term
| What is the data collection error-interaction bias? |
|
Definition
| inaccurate times just from being there-operator speed up or slow down |
|
|
Term
| What is the data collection error-perception |
|
Definition
| measurer seeing what they want to see |
|
|
Term
| What is the data collection error-sampling? |
|
Definition
| Data that doesn't represent the process (not timing night shift) |
|
|
Term
| What is simple random sampling |
|
Definition
| selecting n units from a population so that every n has equal chance of being selected |
|
|
Term
| What is stratified random sampling |
|
Definition
| when population has natural strata, randomly sample within each |
|
|
Term
| What is systematic random sampling? |
|
Definition
| choosing a unit and then the next at evenly spaced intervals until reaching desired amount |
|
|
Term
| cluster or multi-stage random sampling |
|
Definition
| population units aren't available but clusters are (cities, departments) and then sample from each cluster |
|
|
Term
| what is a defective unit counted as? |
|
Definition
| 1 regardless of the number of defects on the until |
|
|
Term
| what are defects per unit of production counted as? |
|
Definition
| multiple defects on any given unit |
|
|
Term
| what does throughput yield measure? |
|
Definition
| each step in the process to determine defects per amount produced to establish a baseline |
|
|
Term
|
Definition
| take the first throughput yield times the second throughput yield and so on |
|
|
Term
|
Definition
| defects per million opportunities |
|
|
Term
|
Definition
| (#defects)/(units X opportunities per unit) X 1,000,000 |
|
|
Term
|
Definition
| A measure of potential process capability |
|
|
Term
|
Definition
| Measures actual process capability |
|
|
Term
| What is first pass yield? |
|
Definition
| units that meet spec without any rework |
|
|
Term
| why is first pass yield not great to use? |
|
Definition
| rework is often hidden and it does not capture multiple defects on the same unit |
|
|
Term
|
Definition
parts per million defective
#defects/#units X 1,000,000 |
|
|
