VARIABLE MSA - GAUGE R & R - SETUP

Instructions: Information Entry Key

1) Enter information in Setup worksheet > setup information (op
2) Enter information in Data worksheet > required information
3) Remove any data in cells highlighted yellow > excess information - c
4) Review Results worksheet > results information
5) Review Bias & Consistency charts > intermediate calculati
6) Review Averages & Ranges charts

SETUP

Date Units Prepared by

Gage Name Characteristic Appraiser A Name

Gage Type Specification Appraiser B Name

Gage Number Test Number Appraiser C Name

Sample Size (p) 10 Appraisers (o) 3 Trials (n) 3 Upper Specification

Lower Specification

Tolerance (Upper Spec Limit - Lower Spec

Next step: Enter Data

SETUP
n Entry Key
> setup information (optional)
> required information
> excess information - clear contents
> results information
> intermediate calculations/results

Upper Specification Limit

Lower Specification Limit

Spec Limit - Lower Spec Limit) 0.00

 VARIABLE MSA - GAUGE R & R - DATA ENTRY
Enter data in gold cells!
Appraiser (A):
Trial\Sample 1 2 3 4 5 6 7 8
1
2
3
Average - - - - - - - -
Range - - - - - - - -

Appraiser (B):
Trial\Sample 1 2 3 4 5 6 7 8
1
2
3
Average - - - - - - - -
Range - - - - - - - -

Appraiser (C):
Trial\Sample 1 2 3 4 5 6 7 8
1
2
3
Average - - - - - - - -
Range - - - - - - - -

Part Average - - - - - - - -

Previous step: Setup Part Variance (sp2ave) 0.00 Range of Part Avera
Next step: View Results Operator Variance (s 2
o ave ) 0.00 Range of Appraiser Avera
DATA ENTRY

9 10 Average
0.00
0.00
0.00
- - Xbar(A) 0.00
- - Rbar(A) 0.00

9 10 Average
0.00
0.00
0.00
- - Xbar(B) 0.00
- - Rbar(B) 0.00

9 10 Average
0.00
0.00
0.00
- - Xbar(C) 0.00
- - Rbar(C) 0.00

- - Xdoublebar 0.00
Rdoublebar 0.00
Range of Part Averages Rp 0.00
Range of Appraiser Averages Ro 0.00
 VARIABLE MSA - GAUGE R & R - RESULTS
Click here for results interpretation
Click here for a key to terms used in the equations
Component
AIAG Meth
MoreSteam Recommended Alternative Variance Method
% of Total Variation % of Total Variation
REPEATABILITY - EQUIPMENT VARIATION (EV)
EV = Rdoublebar / d2*EV %EV = EV2/TV2 %EV =
EV = 0 %EV = %EV =

REPRODUCIBILITY - APPRAISER VARIATION (AV)

AV = {[Ro / d2*AV]2 - [EV2 / (n x p)]}0.5 %AV = AV2/TV2 %AV =
AV = 0 %AV = %AV =

REPEATABILITY & REPRODUCIBILITY (GRR)

GRR = (EV2 + AV2)0.5 %GRR = GRR2/TV2 %GRR =
GRR = 0 %GRR = %GRR =

PRODUCT VARIATION (PV)

PV = Rp / d2*PV %PV = PV2/TV2 %PV =
PV = 0 %PV = %PV =

S = 100% S <=> 100%

TOTAL VARIATION (TV)
TV = (EV2 + AV2 + PV2)0.5
Review Bias & Consistency Charts
TV = 0
Component Variance
Method

Interpretation of Results for the Component Variance Method

Relative Utility: Goal - at least 80% of measurement is error attributable to the product
Range of % PV Actual %PV This Means:
>= 80%
50% <= %PV < 80%
20% <= %PV < 50%
< 20%

Bias: Goal - all appraiser averages within Bias Chart control limits
Value Result: This Means:
0
All appraisers averages are within the
1 or 2* There is no appraiser bias present
Bisas Chart control limits
# appraisers

Consistency: Goal - all appraisers ranges within Consistency Chart control limits
Value Result: This Means:
0
All appraisers ranges are within the The results are consistent across
1 or 2*
Consistency Chart control limits appraisers
 All appraisers ranges are within the The results are consistent across
Consistency Chart control limits appraisers
# appraisers
* 1 if there are 2 appraisers, 2 if there are 3 appraisers
Back to top

Interpretation of Results for the AIAG Method

Relative Utility: Goal - gauge R&R% is less than 10%

Range of % GRR Actual GRR% This Means:
<= 10%
10% < %GRR <= 30%
> 30%

System Resolution: Goal - all appraisers have at least 50% of average values outside the Averages Chart control limits
Value Result: This Means:
# Appraisers 3 appraisers have less than 50% of the The measurement system may not
averages outside the Averages Chart have adequate resolution to detect
< # Appraisers control limits part-to-part variation

Consistency: Goal - all appraisers have all range values below the Range Chart upper control limit
Value Result: This Means:
0 All appraisers ranges are within the Range The results are consistent across
# appraisers Chart control limits appraisers

* 1 if there are 2 appraisers, 2 if there are 3 appraisers

Back to top
Key to Abbreviations Used in Formulas
Rdoublebar = grand average of all Appraiser ranges
d2* = a constant used to estimate Standard Deviation
Ro = Range of Appraiser averages
Rp = Range of Part averages
p = Sample Size
n = Number of Trials
RESULTS
Previous step: Enter Data
AIAG Method
% of Total Variation % of Tolerance

EV/TV %EV = EV/Tol

%EV = #N/A

AV/TV %AV = AV/Tol

%AV = #N/A

GRR/TV %GRR = GRR/Tol

%GRR = #N/A

PV/TV %PV = PV/Tol

%PV = #N/A

S <=> 100%

Review Averages & Range Charts

AIAG Method

Refer To:
AppraiserBiasChart

Refer To:
AppraiserConsistencyChart
 AppraiserConsistencyChart

ges Chart control limits

Refer To:

AveragesChart

Refer To:

RangeChart
 VARIABLE MSA - GAUGE R & R - BIAS & CONSISTENCY CHARTS
Click here for the Appraiser Consistency Chart Click here for the Control Limits Calculations

Appraiser Bias

0.9

0.8

0.7

0.6
Average

0.5

0.4

0.3

0.2

0.1 UCL
LCL
0
0 0 0

Appraiser

Appraiser Consistency

0.9

0.8

0.7

0.6
Range

0.5

0.4

0.3

0.2
UCL
0.1
LCL

0
 Rang
0.5

0.4

0.3

0.2
UCL
0.1
LCL

0
0 0 0

Appraiser

Back to top
CONTROL LIMIT CALCULATIONS

Appraiser Bias Chart Appraiser Consistency Chart

Xdoublebar + ( Rdoublebar X ANOME) = UCL ( Rdoublebar X
0.0000 + 0.0000 X 0.192 = 0.000 0.0000 X

Xdoublebar - ( Rdoublebar X ANOME) = LCL ( Rdoublebar X

0.0000 - 0.0000 X 0.192 = 0.000 0.0000 X

Appraiser Chart Critical Value Tables

Back to top
STENCY CHARTS
Back to the Results
If all data points are
within the control
limits, there is a
95% probability
that there is no
operator bias
present.

If all data points are

within the control
limits, there is a
95% probability
that there is no
inconsistency of
measurement
accross
appraisers.
 0

Appraiser Consistency Chart

UMR) = UCL
1.315 = 0.000

LMR) = LCL
0.669 = 0.000
 VARIABLE MSA - GAUGE R & R - AVERAGE AND RANGE CHARTS
Click here for the Range Chart Click here for the Control Limits Calculations

Averages Chart
1.000

0.900

0.800

0.700

0.600
Average

0.500

0.400

0.300

0.200

0.100

0.000
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9

Trial Number
Series1 Series2 Series3 UCL
Mean LCL

Range Chart
1.000

0.900

0.800

0.700

0.600
Range

0.500

0.400

0.300

0.200
 0.600

Range
0.500

0.400

0.300

0.200

0.100

0.000
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9

Trial Number
Series1 Series2 Series3
UCL Mean

Back to top
CONTROL CHART CONTROL LIMIT CALCULATIONS

Range Variation # Trials D4

Rbar(A) 0.0000 2 3.267
Rbar(B) 0.0000 3 2.575
Rbar(C) 0.0000
Sum 0.0000 # Trials A2 Xdoublebar +
Rdoublebar 0.0000 2 1.880 0.0000 +
3 1.023
Max X Bar 0.0000 Xdoublebar -
Min X Bar 0.0000 0.0000 -
X bar Diff 0.0000

Back to top
NGE CHARTS
Click here to go back to the Results
If 50% or more of the A
averages for each appraiser ###
are are outside the control ###
limits, the measurement ###
system has adequate
###
resolution to detect part-to-
part variaion. ###
###
###
###
###
###
###
###
###
###
###
###
###
###
###
###
###
###
###
###
2 3 4 5 6 7 8 9 10 ###
###
###
UCL
###
###
If all data points are within ###
the control limit, there is no ###
inconsistency of ###
measurement accross ###
appraisers.
2 3 4 5 6 7 8 9 10

ALCULATIONS

Rdoublebar x D4 = UCLR
0.0000 X 2.575 = 0.000
Note: LCL is zero with < 8 trials

( Rdoublebar X A2) = UCLX

0.0000 X 1.023 = 0.000

( Rdoublebar X A2) = LCLX

0.0000 X 1.023 = 0.000
 VARIABLE MSA - GAUGE R & R - BIAS & CONSISTENCY CHART CRITICAL VALUES

Appraiser Bias Appraiser Consistency

Analysis of Mean Effects (ANOME) Analysis of Mean Range (ANOMR)
LMR UMR
n n n
k m 2 3 ANOME.05 k m 2 3 LMR.05 k m 2
6 2 0.610 0.299 0.192 6 2 0.395 0.575 0.669 6 2 1.605
6 3 1.084 0.519 6 3 0.136 0.333 6 3 2.133
8 2 0.501 0.253 8 2 0.475 0.635 8 2 1.525
9 3 0.814 0.408 9 3 0.246 0.442 9 3 1.915
10 2 0.435 0.224 10 2 0.530 0.672 10 2 1.470
12 2 0.389 0.203 12 2 0.569 0.703 12 2 1.431
12 3 0.678 0.346 12 3 0.329 0.511 12 3 1.784
14 2 0.357 0.186 14 2 0.603 0.724 14 2 1.397
15 3 0.592 0.306 15 3 0.388 0.559 15 3 1.701
16 2 0.331 0.172 16 2 0.630 0.743 16 2 1.370
18 2 0.309 0.163 18 2 0.649 0.757 18 2 1.351
18 3 0.531 0.278 18 3 0.436 0.599 18 3 1.637
20 2 0.292 0.140 20 2 0.668 0.781 20 2 1.332
21 3 0.485 0.245 21 3 0.478 0.638 21 3 1.585
24 2 0.264 0.123 24 2 0.696 0.803 24 2 1.304
24 3 0.451 0.226 24 3 0.505 0.658 24 3 1.547
25 3 0.416 0.204 25 3 0.322 0.506 25 3 1.861
27 3 0.406 0.208 27 3 0.530 0.681 27 3 1.515
28 2 0.242 0.102 28 2 0.719 0.827 28 2 1.218
30 2 0.233 0.099 30 2 0.728 0.831 30 2 1.272
30 3 0.400 0.192 30 3 0.556 0.669 30 3 1.485

m - # appraisers, n = number of trials, k = # of measurements per trial (m x n)

In the absence of systematic differences between these m Averages/Ranges, they will all fall within the calculated
control limits 95% of the time.

These tables adopted from EMP III Using Imperfect Data, Donald Wheeler, 2006, SPC Press
CRITICAL VALUES

nsistency
ange (ANOMR)
UMR
n
3 UMR.05
1.425 1.315
1.775
1.365
1.625
1.328
1.297
1.534
1.276
1.476
1.257
1.243
1.436
1.219
1.389
1.197
1.360
1.562
1.333
1.173
1.169
1.315

l within the calculated

 VARIABLE MSA - GAUGE R & R - d2* TABLE
m
2 3 4 5 6 7 8 9
1 1.41 1.91 2.24 2.48 2.67 2.83 2.96 3.08
2 1.28 1.81 2.15 2.4 2.6 2.77 2.91 3.02
3 1.23 1.77 2.12 2.38 2.58 2.75 2.89 3.01
4 1.21 1.75 2.11 2.37 2.57 2.74 2.88 3
5 1.19 1.74 2.1 2.36 2.56 2.73 2.87 2.99
6 1.18 1.73 2.09 2.35 2.56 2.73 2.87 2.99
7 1.17 1.73 2.09 2.35 2.55 2.72 2.87 2.99
8 1.17 1.72 2.08 2.35 2.55 2.72 2.87 2.98
g

9 1.16 1.72 2.08 2.34 2.55 2.72 2.86 2.98

10 1.16 1.72 2.08 2.34 2.55 2.72 2.86 2.98
11 1.16 1.71 2.08 2.34 2.55 2.72 2.86 2.98
12 1.15 1.71 2.07 2.34 2.55 2.72 2.85 2.98
13 1.15 1.71 2.07 2.34 2.55 2.71 2.85 2.98
14 1.15 1.71 2.07 2.34 2.54 2.71 2.85 2.98
15 1.15 1.71 2.07 2.34 2.54 2.71 2.85 2.98
> 15 1.128 1.693 2.059 2.326 2.534 2.704 2.847 2.97

d2*EV
Equipment 1.693 for EV --> m =# Trials, g = # Parts x # Appraisers

d2*AV
Appraisers 1.91 for AV --> m = # Appraisers, g = 1 (only 1 range)

d2*PV
Parts 3.18 for PV --> m = # Parts, g = 1 (only 1 range)
 10
3.18
3.13
3.11
3.1
3.1
3.1
3.1
3.09
3.09
3.09
3.09
3.09
3.09
3.08
3.08
3.078