PRECISION STATEMENTS IN

UOP METHODS
UOP Method 666-82
SCOPE
This method is for developing precision statements as reported in UOP Methods. Precision data in all
new methods or in methods with a 67 suffix or later were developed by the described procedure.

OUTLINE OF METHOD
Using a specified method of test, either a minimum of 5 duplicate sample analyses or a minimum of 5
replicate analyses are performed. Results are tabulated and the estimated standard deviation (esd) and
allowable difference are calculated.

PROCEDURE
The laboratory supervisor under whose jurisdiction the method was developed is responsible for
collection of the necessary precision data. All data should be obtained by the same operator. It should
preferably be of the in-line sample, duplicate result type but may also be developed from replicates on the
same sample.
Obtain analyses on a minimum of 5 duplicate samples (pairs) or 5 replicates of the sample. Six or more
pieces of data are preferable since the calculation will be more precise. When a method claims applicability
to a broad concentration range or to different sample types, several esds should be determined to fully
cover the entire range. Calculate an estimated standard deviation and an allowable difference at the 95%
probability level as shown in CALCULATIONS.

CALCULATIONS
A. Estimated Standard Deviation (esd) from Pairs (Mean Range)
Calculate the average difference (mean range) as follows:

Average difference =

A
B

where:

A = difference between two results comprising a pair

IT IS THE USERS RESPONSIBILITY TO ESTABLISH APPROPRIATE PRECAUTIONARY PRACTICES AND TO
DETERMINE THE APPLICABILITY OF REGULATORY LIMITATIONS PRIOR TO USE. EFFECTIVE HEALTH AND
SAFETY PRACTICES ARE TO BE FOLLOWED WHEN UTILIZING THIS PROCEDURE. FAILURE TO UTILIZE THIS
PROCEDURE IN THE MANNER PRESCRIBED HEREIN CAN BE HAZARDOUS. MATERIAL SAFETY DATA SHEETS
(MSDS) OR EXPERIMENTAL MATERIAL SAFETY DATA SHEETS (EMSDS) FOR ALL OF THE MATERIALS USED IN
THIS PROCEDURE SHOULD BE REVIEWED FOR SELECTION OF THE APPROPRIATE PERSONAL PROTECTION
EQUIPMENT (PPE).
COPYRIGHT 1982 UOP LLC
ALL RIGHTS RESERVED

UOP Methods are available through ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken PA 19428-2959,
United States. The Methods may be obtained through the ASTM website, www.astm.org, or by contacting Customer Service at
[email protected], 610.832.9555 FAX, or 610.832.9585 PHONE.
--`,``,``,,,,``,,,,`,,,,,``,,,,-`-`,,`,,`,`,,`---

Copyright UOP
Provided by IHS under license with UOP
No reproduction or networking permitted without license from IHS

Licensee=Ameriven/5953923001
Not for Resale, 06/06/2007 14:47:59 MDT

--`,``,``,,,,``,,,,`,,,,,``,,,,-`-`,,`,,`,`,,`---

2 of 5

B = number of pairs
Calculate the esd as follows:

esd from pairs =

C
d

where:

C = average difference, defined above

d = divisor, obtained from Table 1

Example Calculation
Sample Data, Percent Platinum
Sample
No.

Trial 1

Trial 2

Difference

1
2
3
4
5

0.3824
0.3880
0.4075
0.3905
0.3780

0.3837
0.3865
0.4085
0.3915
0.3800

0.0013
0.0015
0.0010
0.0010
0.0020

1. Calculate the average difference (Mean Range) from the data.

0.0013 + 0.0015 + 0.0010 + 0.0010 + 0.0020

= 0.00136
5
2. Find the appropriate d (5 pairs of data) from Table 1. Divide the average difference by d to obtain
the esd.

esd =

0.00136
= 0.0011% Pt
1.19
Table 1

Divisors to Calculate esd from Mean Range

Number of
Pairs of Data

Divisors,
d

Degrees of
Freedom, n

5
6
7
8
9

1.19
1.18
1.17
1.17
1.16

4.6
5.5
6.4
7.2
8.1

B. Estimated Standard Deviation from Replicates

Calculate the esd from replicates as follows:

esd from replicates =

D2
N1
666-82

Copyright UOP
Provided by IHS under license with UOP
No reproduction or networking permitted without license from IHS

Licensee=Ameriven/5953923001
Not for Resale, 06/06/2007 14:47:59 MDT

3 of 5

where:

D = deviation of an individual value from the average of all values

N = number of replicates

Example Calculation
Sample Data
Trial

% Pt

1
2
3
4
5
6

0.3824
0.3837
0.3850
0.3820
0.3835
0.3840

1. Calculate the average value from the data.

Average =

0.3824 + 0.3837 + 0.3850 + 0.3820 + 0.3835 + 0.3840

= 0.3834
6

2. Calculate the deviation of each value from the average.

Trial

Deviation

1
2
3
4
5
6

0.3834 - 0.3824 = 0.0010

0.3837 - 0.3834 = 0.0003
0.3850 - 0.3834 = 0.0016
0.3834 - 0.3820 = 0.0014
0.3835 - 0.3834 = 0.0001
0.3840 - 0.3834 = 0.0006

3. Calculate the sum of the squares of each deviation.

Trial

Square

1
2
3
4
5
6

(0.0010) (0.0010) = 0.00000100

(0.0003) (0.0003) = 0.00000009
(0.0016) (0.0016) = 0.00000256
(0.0014) (0.0014) = 0.00000196
(0.0001) (0.0001) = 0.00000001
(0.0006) (0.0006) = 0.00000036
Sum of the Squares = 0.00000598

4. Calculate the esd by substituting into the above equation

esd =

0.00000598
=
5

5.98 10 6
= 1.196 10 6 = 1.09 10 3 = 0.0011% Pt
5

C. Allowable Difference (95% Probability)

666-82
--`,``,``,,,,``,,,,`,,,,,``,,,,-`-`,,`,,`,`,,`---

Copyright UOP
Provided by IHS under license with UOP
No reproduction or networking permitted without license from IHS

Licensee=Ameriven/5953923001
Not for Resale, 06/06/2007 14:47:59 MDT

4 of 5

Calculate the allowable difference (95% probability) from the esd obtained from either pairs or replicates
using the appropriate multiplier from Table 2.

Table 2
Multiplier to Calculate Allowable Difference
(95% Probability)
Degrees of
Freedom, n*
4
4.6
5
5.5
6
6.4
7
7.2
8
8.1
9

Multiplier

Degrees of
Freedom, n*

Multiplier

3.92
3.75
3.64
3.55
3.46
3.41
3.34
3.32
3.26
3.25
3.20

10
11
12
13
14
15
20
30
40
120

3.15
3.11
3.08
3.05
3.03
3.01
2.95
2.89
2.86
2.80
2.77

*n = Number of replicates less one, where replicates are used to calculate the esd. Table 1
gives n for determining the multiplier where pairs are used.

Example Calculations
If the esd calculated from 5 pairs = 0.0011% Pt and n from Table 1 = 4.6, then the multiplier from
Table 2 is 3.75 and the allowable difference (95% probability) = (0.0011) (3.75) = 0.0041% Pt.
If the esd calculated from 6 replicates = 0.0011% Pt and n from Table 2 = 5, then the multiplier from
Table 2 is 3.64 and the allowable difference (95% probability) = (0.0011) (3.64) = 0.0040% Pt.

REPORT
Two statements are included in the precision statement of the method; the first will state the calculated
esd and the second, the allowable difference at the 95% probability level. For example,
Based on 6 replicate determinations (or, if pairs were used, based on 5 pairs of analyses), the
estimated standard deviation (esd) for platinum at the 0.38 wt-% level was calculated to be 0.0011%.
Duplicate results by the same operator should not differ by more than 0.0040 wt-% (95% probability)
at the stated level.
When more than one esd is required due to different sample type or concentration ranges, the data should
be presented in a table format. For example,

666-82
--`,``,``,,,,``,,,,`,,,,,``,,,,-`-`,,`,,`,`,,`---

Copyright UOP
Provided by IHS under license with UOP
No reproduction or networking permitted without license from IHS

Licensee=Ameriven/5953923001
Not for Resale, 06/06/2007 14:47:59 MDT

5 of 5

Based on 6 replicate determinations, the estimated standard deviation (esd) was calculated for the
impurities listed in the table below. Duplicate results by the same operator should not differ by more than
the allowable difference shown (95% probability).

Component
Benzene
Toluene

Level,
wt-ppm

esd,
wt-ppm

Allowable
Difference,
wt-ppm

88
27

1.1
0.3

4.0
1.1

NOTES
1. If it is absolutely impossible to develop precision data, a statement such as the following should be
included:
An estimated standard deviation is not reported since insufficient data are available at present to permit
this calculation with at least 4 degrees of freedom.
Such a statement should be accompanied by an estimate such as:
It is estimated that duplicate analyses should agree within 0.2 wt-%.
2. The calculations described in this method provide a qualitative estimate as to the precision of the
method, however, the results obtained from these computations should not be considered exact statistical
statements.
3. Precision data are only an indication of how well the method can be repeated. There is no direct
relationship between repeatability and accuracy.

666-82
--`,``,``,,,,``,,,,`,,,,,``,,,,-`-`,,`,,`,`,,`---

Copyright UOP
Provided by IHS under license with UOP
No reproduction or networking permitted without license from IHS

Licensee=Ameriven/5953923001
Not for Resale, 06/06/2007 14:47:59 MDT