Attribute Gage R & R Effectiveness

Instructions:
1) 2) 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. 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. YOU MUST ENTER THE INFORMATION IN 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. If you or an expert has selected samples to be evaluated and you know what attributes these samples are, enter this information in theAttribute sample 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. 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. To print a copy of the report click on the Print Report icon. To delete the data in the spreadsheet, click on the Delete Data icon. To delete all and begin a new test, click on the Delete All icon 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 All icon to delete all data to begin entering your own data.

3)

4)

5) 6) 7) 8)

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, at best a lower 95% limit of 96.4% can be calculated. Thus, we would have to say that an inspector could be as bad as 96% efficient, even though he/she missed no calls. 100 100 100.0% 100.0% 96.4% < Try out different combinations of number of samples and number of matches < to see the effects of sample size. In this case, 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.

Data Entry

Page 1

Report

Attribute Gage R & R Effectiveness

SCORING REPORT
Attribute Legend
1 pass 2 fail

DATE: NAME: PRODUCT: BUSINESS:

3/10/2006 Our Employee Widgets Assembly

All operators agree within and All Operators between each Other agree with standard

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

Operator #1 Try #1 Try #2

pass pass fail fail fail pass fail pass pass pass pass pass fail fail pass pass fail fail fail pass fail pass pass pass pass pass fail fail -

Operator #2 Try #1 Try #2

pass pass fail fail pass pass fail pass pass fail pass pass fail pass pass pass pass fail fail pass fail pass pass fail pass pass fail fail -

Operator #3 Try #1 Try #2

fail fail fail fail fail pass fail pass pass fail pass pass fail fail fail fail fail fail fail pass fail pass pass fail pass pass fail fail -

Y/N Agree N N N Y N Y Y Y Y N Y Y Y N

Y/N Agree N N N Y N Y N Y N N Y Y Y N

(1) % APPRAISER SCORE ->

100.00% 78.57%

78.57% 64.29%

100.00% 71.43%

% SCORE VS. ATTRIBUTE(2) ->

57.14% SCREEN % EFFECTIVE SCORE vs. ATTRIBUTE(4) -> 42.86%

SCREEN % EFFECTIVE SCORE(3) ->

Note:
(1) (2)

Operator agrees with him/herself on both trials 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
Page 1

Statistical Report - Attribute Gage R&R Study

DATE: NAME: PRODUCT: BUSINESS: 3/10/2006 Our Employee Widgets Assembly %Score vs Attribute2 Operator #1 Operator #2 Operator #3 14 14 14 11 9 10

Source Total Inspected # Matched False Negative (operator rejected good product) False Positive (operator accepted bad product) Mixed 95% UCL 100.0% 95.3% Calculated Score 100.0% 78.6% 95% LCL 76.8% 49.2%

% Appraiser1 Operator #1 Operator #2 Operator #3 14 14 14 14 11 14

100.0% 100.0% 76.8%

1 2 0 95.3% 78.6% 49.2%

1 1 3 87.2% 64.3% 35.1%

3 1 0 91.6% 71.4% 41.9%

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

Screen % Effective Score 14 8

Screen % Effective Score vs Attribute 14 6

82.3% 57.1% 28.9%

71.1% 42.9% 17.7%

% Appraiser
110.0% 100.0% 90.0% 80.0% % Efficiency

95% UCL Calculated Score 95% LCL

% Score vs Appraiser
110.0% 100.0% 90.0% 80.0% % Efficiency 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0%

95% UCL Calculated Score 95% LCL

70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% Operator #1 Operator #2 Operator #3

Operator #1

Operator #2

Operator #3

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

Calculations

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

Operator #1 Try #1 Try #2 1 1 1 1 2 2 2 2 2 2 1 1 2 2 1 1 1 1 1 1 1 1 1 1 2 2 2 2

Operator #2 Try #1 Try #2 1 1 1 1 2 1 2 2 1 2 1 1 2 2 1 1 1 1 2 2 1 1 1 1 2 2 1 2

Operator #3 Try #1 Try #2 2 2 2 2 2 2 2 2 2 2 1 1 2 2 1 1 1 1 2 2 1 1 1 1 2 2 2 2

Y/N Agree FALSE FALSE FALSE TRUE FALSE TRUE TRUE TRUE TRUE FALSE TRUE TRUE TRUE FALSE

Y/N Agree FALSE FALSE FALSE TRUE FALSE TRUE FALSE TRUE FALSE FALSE TRUE TRUE TRUE FALSE

Page 1

Calculations

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

100.00%

78.57%

100.00%

Page 2