You are on page 1of 6

Attribute Gage R & R Effectiveness

Instructions:
1)

The following spreadsheet is used to calculate an Attribute GR&R Effectiveness, in which up to 100 samples
can be evaluated, using 2 or 3 operators.

2)

In the Data Entry worksheet, fill in the appropriate information in the Scoring Report section and enter the type
of Attributes you are evaluating in the Attribute Legend section. THE INFORMATION MUST BE
ENTERED INTO THE ATTRIBUTE LEGEND SECTION OR THE SPREADSHEET WILL NOT
WORK. The attributes can be either alpha or numeric, e.g. Yes, No; pass, fail; go, stop; or 1, 2. You must be
consistent throughout the form and spell properly.

3)

If you or an expert has selected samples to be evaluated and you know what attributes these samples are (Good
vs Bad), enter this information in the STANDARD column. This will enable you to determine how well each
operator can evaluate a set of samples against a known standard. You do not need to enter information in this
column for the spreadsheet to work, although you will not be able to assess the operators against known
standards.

4)

You do not have to specify how many operators or the # of samples that you will be evaluating during the test.
Simply enter the data into the spreadsheet under the specific operator. Remember the attributes must be spelled
properly or the spreadsheet will not analyze the data correctly.

5)

To print a copy of the report click on the Print Report icon.

6)

To delete the data in the spreadsheet, click on the Delete Data icon.

7)

To see a Demo of the Attribute GR&R Effectiveness spreadsheet, click on the Demo icon. Move around the
spread sheet to see the data. When you are finished, click the Delete Data icon to delete all data to begin
entering your own data.

NOTES:

The 95% UCL and 95% LCL represent the 95% upper and lower confidence limits on the
binomial distribution. The Calculated Score is the basic computation reported on the report
page for % Appraiser and % Score vs Attribute. The 95% confidence interval represents the
range within which the true Calculated Score lies given the uncertainty associated with limited
sample sizes. As sample size increases (in this case, Total Inspected) the confidence interval
will get smaller and smaller which indicates more reliable estimates of the true percentages.
In the case of the Demo data, the true Calculated score for Operator 1 could be as low as
76.8% given that only 14 samples inspected, even though there was a 100% Appraiser value
calculated. Also, even though Operator 2 had a lower score, Operators 1 and 3 cannot be
distinguished from Operator 2 because the calculated score of #2 (78.6%) lies within the
confidence limits for Operators 1 and 3.

With a worksheet limitation of 100 samples, the best the lower 95% limit can be is 96.4%.
Thus, we would have to say that the best an inspector could be is 96% efficient; even though
they did not make any mistakes.

Try different combinations of number of samples and number of matches to see the effects of
sample size. EXAMPLE: a sample size of 30 with one non-match will yield a 17%
confidence interval. In order to get reasonable reliability in estimates of efficiency, large
sample sizes will be required.

Attribute Legend5
(Must Enter Information)

DATE:
NAME:

PRODUCT:

BUSINESS:

Optional: Enter Operator Name or use Default


Operator #1

Operator #2

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33

Standard

Try #1

Try #2

Try #1

Try #2

Y/N

Y/N

Agree

Agree

Operator #3

Known Population
Sample
#

All Operators agree with


standard

SCORING REPORT

All operators agree witin


and between each other

Attribute Gage R & R Effectiveness

Try #1

Try #2

34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78

79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
% APPRAISER SCORE(1) ->

#DIV/0!

#DIV/0!

0%

% SCORE VS. ATTRIBUTE(2) ->

Known

Known

Known

SCREEN % EFFECTIVE SCORE(3) ->

0.00%

SCREEN % EFFECTIVE SCORE vs. ATTRIBUTE


Note:
(1) Operator agrees with him/herself on both trials
(2) Operator agrees on both trials with the known standard
(3) All operators agreed within and between themselves
(4) All operators agreed within and between themselves AND agreed with the known standard
(5) Enter Pass/Fail, Good/Bad, Accept/Reject or other labels which indicate status of inspection

(4)

->

0.00%

Statistical Report - Attribute Gage R&R Study


DATE:

30-Dec-1899

NAME:

PRODUCT: 0

BUSINESS: 0

% Appraiser to Self1
% Appraiser Vs Standard2
Operator Operator Operator Operator Operator Operator
#1
#2
#3
#1
#2
#3
0
0
0
0
0
0
0
0
0
0
0
0
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!

Source
Total Inspected
# Matched
95% UCL
Calculated Score
95% LCL
False Negative (operator biased toward rejection) Std = Pass
False Positive (operator biased toward acceptance) Std = Fail
Mixed (Operator accepts and rejects the same part)

Total Inspected
# in Agreement
95% UCL
Calculated Score
95% LCL

0
0
0

Screen % Effective
Score3
100
0
3.6%
0.0%
0.0%

0
0
0

0
0
0

Screen % Effective
Score vs Standard4
100
0
3.6%
0.0%
0.0%

% Apprraiser to Self

% Appraise r Vs Standard
95% LCL

100.0%

100.0%

90.0%

90.0%

80.0%

80.0%

70.0%

70.0%

60.0%

60.0%

% Efficiency

% Efficiency

95% LCL

50.0%
40.0%

50.0%
40.0%

30.0%

30.0%

20.0%

20.0%

10.0%

10.0%

0.0%

0.0%
1

Notes
1) Operator agrees with him/herself on both trials
2) Operator agrees on both trials with the known standard
3) All operators agreed within and between themselves
4) All operators agreed within & between themselves AND agreed with the known standard