An American National Standard

Designation: D 6299 – 02

Standard Practice for

Applying Statistical Quality Assurance Techniques to
Evaluate Analytical Measurement System Performance1
This standard is issued under the fixed designation D 6299; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon (e) indicates an editorial change since the last revision or reapproval.

1. Scope D 3764 Practice for Validation of Process Stream Analyzer

1.1 This practice provides information for the design and Systems2
operation of a program to monitor and control ongoing stability D 5191 Test Method for Vapor Pressure of Petroleum Prod-
and precision and bias performance of selected analytical ucts (Mini Method)2
measurement systems using a collection of generally accepted E 177 Practice for Use of the Terms Precision and Bias in
statistical quality control (SQC) procedures and tools. ASTM Test Methods3
E 178 Practice for Dealing with Outlying Observations3
NOTE 1—A complete list of criteria for selecting measurement systems E 456 Terminology Relating to Quality and Statistics3
to which this practice should be applied and for determining the frequency
at which it should be applied is beyond the scope of this practice.
E 691 Practice for Conducting an Interlaboratory Study to
However, some factors to be considered include (1) frequency of use of Determine the Precision of a Test Method3
the analytical measurement system, (2) criticality of the parameter being
measured, (3) system stability and precision performance based on 3. Terminology
historical data, (4) business economics, and (5) regulatory, contractual, or 3.1 Definitions:
test method requirements. 3.1.1 accepted reference value, n—a value that serves as an
1.2 This practice is applicable to stable analytical measure- agreed-upon reference for comparison and that is derived as (1)
ment systems that produce results on a continuous numerical a theoretical or established value, based on scientific principles,
scale. (2) an assigned value, based on experimental work of some
1.3 This practice is applicable to laboratory test methods. national or international organization, such as the U.S. Na-
1.4 This practice is applicable to validated process stream tional Institute of Standards and Technology (NIST), or (3) a
analyzers. consensus value, based on collaborative experimental work
under the auspices of a scientific or engineering group.
NOTE 2—For validation of univariate process stream analyzers, see also
Practice D 3764.
(E 456/E 177)
3.1.2 accuracy, n—the closeness of agreement between an
1.5 This practice assumes that the normal (Gaussian) model observed value and an accepted reference value. (E 456/
is adequate for the description and prediction of measurement E 177)
system behavior when it is in a state of statistical control. 3.1.3 assignable cause, n—a factor that contributes to
NOTE 3—For non-Gaussian processes, transformations of test results variation and that is feasible to detect and identify. (E 456)
may permit proper application of these tools. Consult a statistician for 3.1.4 bias, n—a systematic error that contributes to the
further guidance and information. difference between a population mean of the measurements or
1.6 This practice does not address statistical techniques for test results and an accepted reference or true value. (E 456/
comparing two or more analytical measurement systems ap- E 177)
plying different analytical techniques or equipment compo- 3.1.5 control limits, n—limits on a control chart that are
nents that purport to measure the same property(s). used as criteria for signaling the need for action or for judging
whether a set of data does or does not indicate a state of
2. Referenced Documents statistical control. (E 456)
2.1 ASTM Standards: 3.1.6 lot, n—a definite quantity of a product or material
accumulated under conditions that are considered uniform for
sampling purposes. (E 456)
1
This practice is under the jurisdiction of ASTM Committee D02 on Petroleum
3.1.7 precision, n—the closeness of agreement between test
Products and Lubricants and is the direct responsibility of Subcommittee D02.94 on
Quality Assurance and Statistics.
2
Current edition approved June 10, 2002. Published September 2002. Originally Annual Book of ASTM Standards, Vol 05.02.
3
published as D 6299–98. Last previous edition D 6299–00. Annual Book of ASTM Standards, Vol 14.02.

Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States.

1
 D 6299 – 02
results obtained under prescribed conditions. (E 456) variety of petroleum products and lubricants.
3.1.8 repeatability conditions, n—conditions where mutu- 3.2.9 quality control (QC) sample, n—for use in quality
ally independent test results are obtained with the same test assurance programs to determine and monitor the precision and
method in the same laboratory by the same operator with the stability of a measurement system, a stable and homogeneous
same equipment within short intervals of time, using test material having physical or chemical properties, or both,
specimens taken at random from a single sample of material. similar to those of typical samples tested by the analytical
(E 456, E 177) measurement system. The material is properly stored to ensure
3.1.9 reproducibility conditions, n—conditions under which sample integrity, and is available in sufficient quantity for
test results are obtained in different laboratories with the same repeated, long term testing.
test method, using test specimens taken at random from the 3.2.10 site precision (R8), n—the value below which the
same sample of material. (E 456, E 177) absolute difference between two individual test results obtained
3.2 Definitions of Terms Specific to This Standard: under site precision conditions may be expected to occur with
3.2.1 analytical measurement system, n—a collection of one a probability of approximately 0.95 (95 %). It is defined as 2.77
or more components or subsystems, such as samplers, test times the standard deviation of results obtained under site
equipment, instrumentation, display devices, data handlers, precision conditions.
printouts or output transmitters, that is used to determine a 3.2.11 site precision conditions, n—conditions under which
quantitative value of a specific property for an unknown test results are obtained by one or more operators in a single
sample in accordance with a test method. site location practicing the same test method on a single
3.2.1.1 Discussion—An analytical measurement system measurement system which may comprise multiple instru-
may comprise multiple instruments being used for the same ments, using test specimens taken at random from the same
test method. sample of material, over an extended period of time spanning
3.2.2 blind submission, n—submission of a check standard at least a 15 day interval.
or quality control (QC) sample for analysis without revealing 3.2.11.1 Discussion—Site precision conditions should in-
the expected value to the person performing the analysis. clude all sources of variation that are typically encountered
3.2.3 check standard, n—in QC testing, a material having during normal, long term operation of the measurement sys-
an accepted reference value used to determine the accuracy of tem. Thus, all operators who are involved in the routine use of
a measurement system. the measurement system should contribute results to the site
precision determination. If multiple results are obtained within
3.2.3.1 Discussion—A check standard is preferably a mate-
a 24–h period, then it is recommended that the number of
rial that is either a certified reference material with traceability
results used in site precision calculations be increased to
to a nationally recognized body or a material that has an
capture the longer term variation in the system.
accepted reference value established through interlaboratory
3.2.12 site precision standard deviation, n—the standard
testing. For some measurement systems, a pure, single com-
deviation of results obtained under site precision conditions.
ponent material having known value or a simple gravimetric or
3.2.13 validation audit sample, n—a QC sample or check
volumetric mixture of pure components having calculable
standard used to verify precision and bias estimated from
value may serve as a check standard. Users should be aware
routine quality assurance testing.
that for measurement systems that show matrix dependencies,
3.3 Symbols:
accuracy determined from pure compounds or simple mixtures
3.3.1 ARV—accepted reference value.
may not be representative of that achieved on actual samples.
3.3.2 EWMA—exponentially weighted moving average.
3.2.4 common (chance, random) cause, n—for quality as-
3.3.3 I—individual observation (as in I-chart).
surance programs, one of generally numerous factors, individu-
3.3.4 MR—moving range.
ally of relatively small importance, that contributes to varia-
3.3.5 MR —average of moving range.
tion, and that is not feasible to detect and identify.
3.3.6 QC—quality control.
3.2.5 double blind submission, n—submission of a check 3.3.7 R8—site precision.
standard or QC sample for analysis without revealing the check 3.3.8 sR8—site precision standard deviation.
standard or QC sample status and expected value to the person 3.3.9 VA—validation audit.
performing the analysis. 3.3.10 x2—chi squared.
3.2.6 expected value, n—for a QC sample analyzed using an 3.3.11 l—lambda.
in-statistical control measurement system, the estimate of the
theoretical limiting value to which the average of results tends 4. Summary of Practice
when the number of results approaches infinity. 4.1 QC samples and check standards are regularly analyzed
3.2.7 in-statistical-control, adj—a process, analytical mea- by the measurement system. Control charts and other statistical
surement system, or function that exhibits variations that can techniques are presented to screen, plot, and interpret test
only be attributable to common cause. results in accordance with industry-accepted practices to as-
3.2.8 proficiency testing, n—determination of a laboratory’s certain the in-statistical-control status of the measurement
testing capability by participation in an interlaboratory cross- system.
check program. 4.2 Statistical estimates of the measurement system preci-
3.2.8.1 Discussion—ASTM Committee D02 conducts pro- sion and bias are calculated and periodically updated using
ficiency testing among hundreds of laboratories, using a wide accrued data.

2
 D 6299 – 02
4.3 In addition, as part of a separate validation audit analytical measurement system.
procedure, QC samples and check standards may be submitted 6.2.1 A check standard may be a commercial standard
blind or double-blind and randomly to the measurement system reference material when such material is available in appropri-
for routine testing to verify that the calculated precision and ate quantity, quality and composition.
bias are representative of routine measurement system perfor- NOTE 5—Commercial reference material of appropriate composition
mance when there is no prior knowledge of the expected value may not be available for all measurement systems.
or sample status.
6.2.2 Alternatively, a check standard may be prepared from
5. Significance and Use a material that is analyzed under reproducibility conditions by
5.1 This practice can be used to continuously demonstrate multiple measurement systems. The accepted reference value
the proficiency of analytical measurement systems that are (ARV) for this check standard shall be the average after
used for establishing and ensuring the quality of petroleum and statistical examination and outlier treatment has been applied.4
petroleum products. 6.2.2.1 Exchange samples circulated as part of an interlabo-
5.2 Data accrued, using the techniques included in this ratory exchange program, or round robin, may be used as check
practice, provide the ability to monitor analytical measurement standards. For an exchange sample to be usable as a check
system precision and bias. standard, the standard deviation of the interlaboratory ex-
5.3 These data are useful for updating test methods as well change program shall not be statistically greater than the
as for indicating areas of potential measurement system im- reproducibility standard deviation for the test method. An
provement. F-test should be applied to test acceptability.
6. Reference Materials NOTE 6—The uncertainty in the ARV is inversely proportional to the
square root of the number of values in the average. This practice
6.1 QC samples are used to establish and monitor the recommends that a minimum of 16 non-outlier results be used in
precision of the analytical measurement system. calculating the ARV to reduce the uncertainty of the ARV by a factor of
6.1.1 Select a stable and homogeneous material having 4 relative to the measurement system single value precision. The bias tests
physical or chemical properties, or both, similar to those of described in this practice assume that the uncertainty in the ARV is
typical samples tested by the analytical measurement system. negligible relative to the measurement system precision. If less than 16
values are used in calculating the average, this assumption may not be
NOTE 4—When the QC sample is to be utilized for monitoring a valid.
process stream analyzer performance, it is often helpful to supplement the NOTE 7—Examples of exchanges that may be acceptable are ASTM
process analyzer system with a subsystem to automate the extraction, D02.CS92 ILCP program; ASTM D02.01 N.E.G.; ASTM D02.01.A
mixing, storage, and delivery functions associated with the QC sample. Regional Exchanges; International Quality Assurance Exchange Program,
administered by Alberta Research Council.
6.1.2 Estimate the quantity of the material needed for each
specific lot of QC sample to (1) accommodate the number of 6.2.3 For some measurement systems, single, pure compo-
analytical measurement systems for which it is to be used nent materials with known value, or simple gravimetric or
(laboratory test apparatuses as well as process stream analyzer volumetric mixtures of pure components having calculable
systems) and (2) provide determination of QC statistics for a value may serve as a check standard. For example, pure
useful and desirable period of time. solvents, such as 2,2-dimethylbutane, are used as check stan-
6.1.3 Collect the material into a single container and isolate dards for the measurement of Reid vapor pressure by Test
it. Method D 5191. Users should be aware that for measurement
6.1.4 Thoroughly mix the material to ensure homogeneity. systems that show matrix dependencies, accuracy determined
6.1.5 Conduct any testing necessary to ensure that the QC from pure compounds or simple mixtures may not be repre-
sample meets the characteristics for its intended use. sentative of that achieved on actual samples.
6.1.6 Package or store QC samples, or both, as appropriate 6.3 Validation audit (VA) samples are QC samples and
for the specific analytical measurement system to ensure that check standards, which may, at the option of the users, be
all analyses of samples from a given lot are performed on submitted to the measurement system in a blind, or double
essentially identical material. If necessary, split the bulk blind, and random fashion to verify precision and bias esti-
material collected in 6.1.3 into separate and smaller containers mated from routine quality assurance testing.
to help ensure integrity over time. (Warning—Treat the
material appropriately to ensure its stability, integrity, and 7. Quality Assurance (QA) Program for Individual
homogeneity over the time period for which it is to stored and Measurement Systems
used. For samples that are volatile, such as gasoline, storage in 7.1 Overview—A QA program (1)5 can consist of five
one large container that is repeatedly opened and closed can primary activities: (1) monitoring stability and precision
result in loss of light ends. This problem can be avoided by through QC sample testing, (2) monitoring accuracy, (3)
chilling and splitting the bulk sample into smaller containers,
each with a quantity sufficient to conduct the analysis. Simi-
larly, samples prone to oxidation can benefit from splitting the 4
Refer to Research Report RR:D02–1007 and Practices E 178 and E 691 in
bulk sample into smaller containers that can be blanketed with ASTM Standards on Precision and Bias for Various Applications, ASTM Interna-
tional, for guidance in statistical and outlier treatment of data. Request PCN:03-
an inert gas prior to being sealed and leaving them sealed until 512088-34.
the sample is needed.) 5
The boldface numbers in parentheses refer to the list of references at the end of
6.2 Check standards are used to estimate the accuracy of the this standard.

3
 D 6299 – 02
periodic evaluation of system performance in terms of preci- sufficient quantity, then separate QC samples are employed. In this case,
sion or bias, or both, (4) proficiency testing through participa- the accuracy (see 7.3) is monitored less frequently, and the QC sample
tion in interlaboratory exchange programs where such pro- testing (see 7.2) is used to demonstrate the stability of the measurement
system between accuracy tests.
grams are available, and (5) a periodic and independent system
validation using VA samples may be conducted to provide 7.4.3 It is recommended that a QC sample be analyzed at the
additional assurance of the system precision and bias metrics beginning of any set of measurements and immediately after a
established from the primary testing activities. At minimum, change is made to the measurement system.
the QA program must include at least item one. 7.4.4 Establish a protocol for testing so that all persons who
routinely operate the system participate in generating QC test
NOTE 8—For some measurement systems, suitable check standard data.
materials may not exist, and there may be no reasonably available 7.4.5 Handle and test the QC and check standard samples in
exchange programs to generate them. For such systems, there is no means
of verifying the accuracy of the system, and the QA program will only
the same manner and under the same conditions as samples or
involve monitoring stability and precision through QC sample testing. materials routinely analyzed by the analytical measurement
system.
7.2 Monitoring System Stability and Precision Through QC 7.4.6 When practical, randomize the time of check standard
Sample Testing—QC test specimen samples from a specific lot and additional QC sample testing over the normal hours of
are introduced and tested in the analytical measurement system measurement system operation, unless otherwise prescribed in
on a regular basis to establish system performance history in the specific test method.
terms of both stability and precision.
7.3 Monitoring Accuracy: NOTE 11—Avoid special treatment of QC samples designed to get a
better result. Special treatment seriously undermines the integrity of
7.3.1 Check standards can be tested in the analytical mea-
precision estimates.
surement system on a regular basis to establish system perfor-
mance history in terms of accuracy. 7.5 Evaluation of System Performance in Terms of Precision
7.3.2 For measurement systems where calibration is estab- and Bias:
lished by using multiple standards of known values, such as 7.5.1 Pretreat and screen results accumulated from QC and
materials certified by or traceable to the national certification check standard testing. Apply statistical techniques to the
bodies such as NIST, JIS, BSI, and so forth, and where the total pretreated data to identify erroneous data. Plot appropriately
number of standards used exceed the number of parameters pretreated data on control charts.
estimated by the calibration equation, an alternative approach 7.5.2 Periodically analyze results from control charts, ex-
(instead of check standard testing) to infer system accuracy is cluding those data points with assignable causes, to quantify
to compare the statistics associated with the calibration equa- the bias and precision estimates for the measurement system.
tion to previously established measurement system precision 7.6 Proficiency Testing:
and to standard errors of the calibration standards used. 7.6.1 Participation in regularly conducted interlaboratory
Coverage of this type of statistical techniques for accuracy exchanges where typical production samples are tested by
inference is beyond the scope of this practice. Users are multiple measurement systems, using a specified (ASTM) test
advised to enlist the services of a statistician when using this protocol, provide a cost-effective means of assessing measure-
approach to infer system accuracy instead of check standard ment system accuracy relative to average industry perfor-
testing. mance. Such proficiency testing can be used instead of check
standard testing for systems where the timeliness of the
7.4 Test Program Conditions/Frequency:
accuracy check is not critical. Proficiency testing may be used
7.4.1 Conduct both QC sample and check standard testing
as a supplement to accuracy monitoring by way of check
under site precision conditions.
standard testing.
NOTE 9—It is inappropriate to use test data collected under repeatability 7.6.2 Participants plot their signed deviations from the
conditions to estimate the long term precision achievable by the site consensus values (exchange averages) on control charts in the
because the majority of the long term measurement system variance is due same fashion described below for check standards, to ascertain
to common cause variations associated with the combination of time,
operator, reagents, instrumentation calibration factors, and so forth, which
if their measurement processes are non-biased relative to
would not be observable in data obtained under repeatability conditions. industry average.
7.7 Independent System Validation—Periodically, at the dis-
7.4.2 Test the QC and check standard samples on a regular cretion of users, VA samples may be submitted blind or double
schedule, as appropriate. Principal factors to be considered for blind for analysis. Precision and bias estimates calculated using
determining the frequency of testing are (1) frequency of use of VA samples test data can be used as an independent validation
the analytical measurement system, (2) criticality of the pa- of the routine QA program performance statistics.
rameter being measured, (3) established system stability and
precision performance based on historical data, (4) business NOTE 12—For measurement systems susceptible to human influence,
the precision and bias estimates calculated from data where the analyst is
economics, and (5) regulatory, contractual, or test method
aware of the sample status (QC or check standard) or expected values, or
requirements. both, may underestimate the precision and bias achievable under routine
NOTE 10—At the discretion of the laboratory, check standards may be operation. At the discretion of the users, and depending on the criticality
used as QC samples. In this case, the results for the check standards may of these measurement systems, the QA program may include periodic
be used to monitor both stability (see 7.2) and accuracy (see 7.3) blind or double-blind testing of VA samples.
simultaneously. If check standards are expensive, or not available in 7.7.1 The specific design and approach to the VA testing

4
 D 6299 – 02
program will depend on features specific to the measurement where the standard deviation at the ARV level is the published reproduc-
system and organizational requirements, and is beyond the ibility standard deviation. In the event that no published reproducibility
intended scope of this practice. Some possible approaches are exists and the ARV was established through round robin testing, standard
deviations determined from round robin testing may be used.
noted as follows.
7.7.1.1 If all QC samples or check standards, or both, are 8.2.2.3 If there is no published reproducibility standard
submitted blind or double blind and the results are promptly deviation and the ARV was not arrived at by round robin
evaluated, then additional VA sample testing may not be testing, a standard deviation should be determined by users in
necessary. a technically acceptable manner.
7.7.1.2 QC samples or check standards, or both, may be NOTE 14—It is recommended that the method used to determine the
submitted as unknown samples at a specific frequency. Such standard deviation be developed under the guidance of a statistician.
submissions should not be so regular as to compromise their NOTE 15—To calculate the reproducibility standard deviations from
blind status. published reproducibilities, divide the accepted reproducibility value at
7.7.1.3 Retains of previously analyzed samples may be each level by 2.77.
resubmitted as unknown samples under site precision condi- 8.2.3 Pretreatment of results for VA samples is done in the
tions. Generally, data from this approach can only yield same manner as described in 8.2.1 and 8.2.2.
precision estimates as retain samples do not have ARVs. 8.3 Assessment of Initial Results—Assessment techniques
Typically, the differences between the replicate analyses are are applied to test results collected during the startup phase of
plotted on control charts to estimate the precision of the or after significant modifications to a measurement system.
measurement system. If precision is level dependent, the Perform the following assessment after at least 15 pretreated
differences are scaled by the standard deviation of the mea- results have become available. The purpose of this assessment
surement system precision at the level of the average of the two is to ensure that these results are suitable for deployment of
results. control charts (described in A1.4).
8. Procedure for Pretreatment, Assessment, and NOTE 16—These techniques can also be applied as diagnostic tools to
Interpretation of Test Results investigate out-of-control situations.
8.1 Overview—Results accumulated from QC, check stan- 8.3.1 Screen for Suspicious Results—Pretreated results
dard, and VA sample testing are pretreated and screened. should first be visually screened for values that are inconsistent
Statistical techniques are applied to the pretreated data to with the remainder of the data set, such as those that could have
achieve the following objectives: been caused by transcription errors. Those flagged as suspi-
8.1.1 Identify erroneous data, cious should be investigated. Discarding data at this stage must
8.1.2 Assess initial results, be supported by evidence gathered from the investigation. If,
8.1.3 Deploy, interpret and maintain of control charts, and after discarding suspicious pretreated results there are less than
8.1.4 Quantify long term measurement precision and bias. 15 values remaining, collect additional data and start over.
8.3.2 Screen for Unusual Patterns—The next step is to
NOTE 13—Refer to the annex for examples of the application of the examine the pretreated results for non-random patterns such as
techniques that are discussed below and described in Section 9.
continuous trending in either direction, unusual clustering, and
8.2 Pretreatment of Test Results—Assessment, control cycles. One way to do this is to plot the results on a run chart
charting, and evaluation are applied only to appropriately (see A1.3) and examine the plot. If any non-random pattern is
pretreated test results. The purpose of pretreatment is to detected, investigate for and eliminate the root cause(s).
standardize the control chart scales so as to allow for data from Discard the data set and start the procedure again.
multiple check standards to be compared on the same chart. 8.3.3 Test “Normality” Assumption—For measurement sys-
8.2.1 For QC sample test results, no data pretreatment is tems with no prior performance history, or as a diagnostic tool,
typically used since results for different QC samples are it is useful to test that the results from the measurement are
generally not plotted on the same chart. adequately described by a normal distribution. One way to do
8.2.2 For check standard sample test results, two cases this is to use a normal probability plot and the Anderson-
apply, depending on the measurement system precision: Darling Statistic (see A1.4). If the results show obvious
8.2.2.1 Case 1—If either (1) all of the check standard test deviation from normality, then the statistical control charting
results are from one or more lots of check standard material techniques described are not directly applicable to the mea-
having the same ARV(s), or (2) the precision of the measure- surement system.
ment system is constant across levels, then pretreatment
NOTE 17—Transformations may lead to normally distributed data, but
consists of calculating the difference between the test result and
these techniques are outside the scope of this practice.
the ARV:
8.4 Control Charts (1, 2)—Individual (I) and moving range
Pretreated result 5 test result 2 ARV~for the sample! (1)
of two (MR) control charts are the recommended tools for (a)
8.2.2.2 Case 2—Test results are for multiple lots of check routine recording of QC sample and check standard test results,
standards with different ARVs, and the precision of the and (b) immediate assessment of the “in statistical control” (3)
measurement system is known to vary with level, status of the system that generated the data. Optionally, the
@test result 2 ARV~for the sample!# exponentially weighted moving average (EWMA) (4, 5) may be
Pretreated result 5 standard deviation at the ARV level (2) overlaid on the I chart to enhance detection power for small
level shifts.

5
 D 6299 – 02
NOTE 18—The control charts and statistical techniques described in this produced by a measurement system that is in statistical control.
practice are chosen for their simplicity and ease of use. It is not the intent This practice presents two procedures to be selected at the
of this practice to preclude use of other statistically equivalent or more users’ discretion.
advanced techniques, or both.
8.7.2 Procedure 1–Concurrent Testing:
8.4.1 Construction of Control Charts—If no obvious un- 8.7.2.1 Collect and prepare a new batch of QC material
usual patterns are detected from the run charts, and no obvious when the current QC material supply remaining can support no
deviation from normality is detected, proceed with construc- more than 20 analyses.
tion of the control charts 8.7.2.2 Concurrently test and record data for the new
8.4.1.1 MR Chart—Construct an MR plot and examine it for material each time a current QC sample is tested. The result for
unusual patterns. If no unusual patterns are found in the MR the new material is deemed valid if the measurement process
plot, calculate and overlay the control limits on the MR plot to in-control status is validated by the current QC material and
complete the MR chart. control chart.
8.4.1.2 I Chart—Calculate control limits and overlay them 8.7.2.3 Optionally, to provide an early indication of the
on the “run chart” to produce the I chart. status of the new batch of QC material, immediately start a run
8.4.1.3 EWMA Overlay—Optionally, calculate the EWMA chart and an MR plot for the new material. After five valid
values and plot them on the I chart. Calculate the EWMA results become available for the new material, convert the run
control limits and overlay them on the I chart. chart into an I chart with trial control limits by adding a center
8.4.2 Control Chart Deployment—Put these control charts line based on the average of the five results and control limits
into operation by regularly plotting the pretreated test results based on the MR from previous control charts for materials at
on the charts and immediately interpreting the charts. the same nominal level. Set trial control limits for the MR chart
8.5 Control Chart Interpretation: based on limits from previous charts for materials at the same
8.5.1 Apply control chart rules (see A1.5) to determine if the nominal level.
data supports the hypothesis that the measurement system is 8.7.2.4 After a minimum of 15 in-control data points are
under the influence of common causes variation only (in collected on the new material, perform an F test of sample
statistical control). variances for the new data set versus the historical variance
8.5.2 Investigate Out-of-Control Points in Detail—Exclude demonstrated at nominal level of the new material. If the
from further data analysis those associated with assignable outcome of the F test is not significant then the precision
causes, provided the assignable causes are deemed not to be estimate is updated by statistically pooling both sample vari-
part of the normal process. ances. A significant F test should trigger an investigation for
NOTE 19—All data, regardless of in-control or out-of-control status,
root cause(s).
needs to be recorded. 8.7.2.5 Construct new I and MR charts (and optional EWMA
overlay) for this new material as per Section 8, using the
8.6 Scenario 1 for Periodic Updating of Control Charts pooled MR .
Parameters: 8.7.2.6 Switch over to the new I and MR charts upon
8.6.1 Scenario 1 covers (a) control charts for a QC material depletion of current QC material.
where there had been no change in the system, but more data 8.7.3 Procedure 2—Q Procedure (see A1.9) (6):
of the same level has been accrued; or (b) control charts for 8.7.3.1 This procedure is designed to alleviate the need for
check standard pretreated results. concurrent testing of two materials. A priori knowledge of the
8.6.2 When a minimum of 15 new in-control data points measurement process historical standard deviation applicable
becomes available, the precision estimate used to calculate the at the new QC material composition and property level is
control limits can be updated to incorporate the information required.
from this new data. Update calculations that involve pooling of
old and new data sets shall be preceded by an F-test (see A1.8) NOTE 20—It is recommended that this standard deviation estimate be
based on at least 50 data points.
of sample variances for the new data set versus the existing
in-control data set. 8.7.3.2 When the Q procedure is operational (minimum of
8.6.3 If the outcome of the F-test is not significant, then the two data points), it can be used in conjunction with a MR chart
precision estimate is updated by statistically pooling both constructed using the observations to provide QA of the
sample variances. A significant F-test should trigger an inves- measurement process.
tigation for assignable causes.
NOTE 21—The Q procedure is not suitable for monitoring measurement
8.7 Scenario 2 for Periodic Updating of Control Charts system bias relative to an external value. It is designed to monitor the
Parameters: stability of the system mean. When used in conjunction with the MR chart,
8.7.1 Scenario 2 covers control chart for QC materials “in statistical control” status of the measurement system can be ascer-
where an assignable cause change in the system had occurred tained.
due to a change in the level for the QC material. Minor or
major differences may exist between QC material batches. 9. Evaluation of System Performance in Terms of
Since control limit calculations for the I chart require a center Precision and Bias
value established by the measurement system, a special tran- 9.1 Site Precision Estimated from Testing of QC Samples:
sition procedure is required to ensure that the center value for 9.1.1 Estimate the site precision of the measurement system
a new batch of QC material is established using results at the level corresponding to a specific lot of QC sample as

6
 D 6299 – 02
2.46 times the MR from the MR chart for that specific lot. corresponding ARVs. Examine the plot for patterns indicative
R8 5 2.46 3 MR (3) of level-dependent bias.
9.3.2 If there is no discernible pattern, perform the t test as
NOTE 22—The site precision standard deviation (sR8) is estimated from described in 9.2 to determine if the average of all the pretreated
the MR chart as R8/2.77 = MR /1.128.
differences plotted on the I chart is statistically different from
9.1.1.1 Alternatively, R8 many be estimated using the root- zero.
mean-square formula for standard deviation: 9.3.2.1 If the outcome of the t test is that the average is not

Œ
n statistically different from zero, then the bias in the measure-
( ~Ii – Ī!2
i51
ment process is negligible.
sR8 5 n–1 (4) 9.3.2.2 If the outcome of the t test is that the average is
statistically different from zero, then there is evidence that the
R8 5 2.77 3 sR8 (5) measurement system is biased. The bias may be level depen-
dent. However, the statistical methodology for estimating the
9.1.1.2 For estimate of site precision standard deviation bias/level relationship is beyond the scope of this practice.
(sR8) using retain results, first obtain the standard deviation of
9.3.3 If there is a discernible pattern in the plot in 9.3.1, then
differences by applying the root-mean-square formula below to
the measurement system may exhibit a level dependent bias.
the differences between the original and retest results for
The statistical methodology for estimating the bias/level rela-
samples with same nominal property level. If measurement
tionship is beyond the scope of this practice.
process precision is known to be level independent, retest
9.3.4 If a bias is detected in 9.3.2.2, or if the plot in 9.3.3
results from samples with different property levels can be used.
exhibits discernible patterns, investigate for root cause(s).
Divide the standard deviation of differences by 1.414 to obtain
the estimate for site precision standard deviation. (sR8).
10. Validation of System Performance Estimates Using
standard deviation of differences 5 (6) VA Samples

Œ( ~individual difference – average difference!2

total number of differences
10.1 If the users decide to include VA sample testing as part
of their QA program, then they should periodically evaluate the
results obtained on the VA samples. The purpose of the
sR8 5 ~standard deviation of differences! 4 1.414 (7) evaluation is to establish whether the system performance
9.1.2 Compare R8 to published reproducibility of the test estimates described in Section 9 are reasonably applicable to
method at the same level, if available. R8 is expected to be less routinely tested samples.
than or equal to the published value. Use the x2 test described 10.2 VA sample test results should be evaluated indepen-
in A1.7. dently through an internal or external audit system, or both. It
9.2 Measurement System Bias Estimated from Multiple is recommended that the internal audit team not be limited to
Measurements of a Single Check Standard—If a minimum of the operators of the measurement system and their immediate
15 test results is obtained on a single check standard material supervisors.
under site precision conditions, then calculate the average of all 10.3 Insofar as possible, analyze the results obtained on the
the in-control individual differences plotted on the I chart. VA samples separately and in the same manner as those from
Perform a t test (see A1.6) to determine if the average is the routine QC and check standard testing program.
statistically different from zero. 10.4 Using F or t tests, or both (see A1.8 and A1.6),
9.2.1 If the outcome of the ttest is that the average is not statistically compare the system performance estimates ob-
statistically different from zero, then the bias in the measure- tained from the VA sample testing program to the measurement
ment process is negligible. system accuracy and precision estimates from the QC sample
9.2.2 If the outcome of the t test is that the average is testing program.
statistically different from zero, then the best estimate of the 10.5 If the comparison reveals that the two estimates of the
measurement process bias at the level of the check standard is measurement system performance are not statistically equiva-
the average. lent, there is cause for concern that the actual performance of
9.3 Measurement System Bias Estimated from Measure- the measurement system may be significantly worse than
ments of Multiple Check Standards—When using multiple estimated. Investigate thoroughly for the assignable cause(s) of
check standards, determine if there is a relationship between this inconsistency, and eliminate it. Until the causes are
the bias and the measurement level. identified and eliminated, the lab precision estimates of Section
9.3.1 Plot the pretreated results as per Section 8 versus their 9 should be considered suspect.

7
 D 6299 – 02
ANNEX

(Mandatory Information)

A1. STATISTICAL QUALITY CONTROL TOOLS

A1.1 Purpose of this Annex TABLE A1.2 Example of a Sequence of Results from a Single
Check Standard
A1.1.1 The purpose of this annex is to provide guidance to
Check Standard Accepted Difference
practitioners, including worked examples, for the proper ex- Sequence Number
Result Reference Value Result - ARV
ecution of the statistical procedures described in this practice. (Yi) (ARV = Xi) Ii
1 55.3 55.88 -0.58
A1.2 Pretreatment of Test Results (8.1 to 8.2.4) 2 55.8 55.88 -0.08
3 56.3 55.88 0.42
A1.2.1 Throughout this annex, {Yi:i=1. . .n} denotes a 4 56.1 55.88 0.22
sequence of as measured test results. {Ii:i=1. . .n} will signify 5 55.8 55.88 -0.08
6 55.5 55.88 -0.38
a sequence of test results after pretreatment, if necessary. 7 55.3 55.88 -0.58
A1.2.2 If {Yi:i=1. . .n} is a sequence of results from a single 8 55.4 55.88 -0.48
QC sample, then 9 56.6 55.88 0.72
10 56.1 55.88 0.22
Ii 5 Yi (A1.1) 11 55.0 55.88 -0.88
12 55.5 55.88 -0.38
with no pretreatment being required. 13 55.5 55.88 -0.38
A1.2.2.1 An example of a sequence of results, Yi, from a 14 55.2 55.88 -0.68
15 56.5 55.88 0.62
single QC sample is given in Columns 2 and 4 of Table A1.1. 16 55.7 55.88 -0.18
A1.2.3 If {Yi:i=1. . .n} is a sequence of results from a single 17 55.6 55.88 -0.28
check standard, from multiple check standards having nomi- 18 55.2 55.88 -0.68
19 55.7 55.88 -0.18
nally the same ARV, or from multiple check standards having 20 56.1 55.88 0.22
different ARVs where the precision of the measurement system 21 56.3 55.88 0.42
does not vary with level, and if { Xi:i=1. . .n} is the sequence 22 55.2 55.88 -0.68
23 55.4 55.88 -0.48
of corresponding ARVs, then 24 55.4 55.88 -0.48
Ii 5 Yi – Xi (A1.2) 25 55.6 55.88 -0.28

The reproducibility standard deviation of the measurement

process must be essentially the same for all values {Xi}. A1.2.4.1 Table A1.3 shows an example of results for mul-
A1.2.3.1 An example of a sequence of results from a single tiple check standards where the precision of the measurement
check standard is given in Table A1.2. The preprocessed result, system is level dependent.
Ii, is given in Column 4 of Table A1.2.
TABLE A1.3 Example of Results for Multiple Check Standards
A1.2.4 If {Yi} is a sequence of results from different check Where the Precision of the Measurement System Is Level
standards, and if the reproducibility varies with the level of the Dependent
accepted reference values, { Xi}, then Result Preprocessed
Raw ARV Raw
Sequence si Result
Ii 5 ~Yi – Xi!/si (A1.3) Result Yi Xi Difference
Number, i Ii
where si are estimates of the reproducibility standard devia- 1 71.0 71.4 -0.40 1.14 -0.35
tion of the measurement process at levels {Xi}. 2 65.8 64.9 0.90 1.10 0.82
3 70.3 70.2 0.10 1.13 0.09
4 66.2 67.7 -1.50 1.11 -1.35
5 93.8 93.4 0.40 1.26 0.32
TABLE A1.1 Example of a Sequence of Results from a Single QC 6 102.9 104.0 -1.10 1.33 -0.83
Sample 7 102.2 101.8 0.40 1.31 0.30
8 103.2 103.9 -0.70 1.32 -0.53
QC/Check Standard QC/Check Standard
Sequence Number Sequence Number 9 100 99.8 0.20 1.30 0.15
Result Result
i i 10 71.6 71.5 0.10 1.14 0.09
Yi= Ii Yi= Ii
11 76.7 76.4 0.30 1.16 0.26
1 55.3 14 55.2 12 61.2 61.8 -0.60 1.08 -0.56
2 55.8 15 56.5 13 44.1 43.9 0.20 0.98 0.20
3 56.3 16 55.7 14 69.71 69.7 0.01 1.13 0.01
4 56.1 17 55.6 15 59.5 59.19 0.31 1.06 0.29
5 55.8 18 55.2 16 99.63 98.87 0.76 1.30 0.59
6 55.5 19 55.7 17 93.7 95.21 -1.51 1.27 -1.19
7 55.3 20 56.1 18 103.77 103.94 -0.17 1.32 -0.13
8 55.4 21 56.3 19 96.18 96.7 -0.52 1.28 -0.41
9 56.6 22 55.2 20 99.7 100.65 -0.95 1.31 -0.73
10 56.1 23 55.4 21 84.32 84.15 0.17 1.21 0.14
11 55.0 24 55.4 22 83.29 83.75 -0.46 1.21 -0.38
12 55.5 25 55.6 23 65.16 65.93 -0.77 1.10 -0.70
13 55.5 24 68.19 68.0 0.19 1.12 0.17

8
 D 6299 – 02
A1.3 The Run Chart
A1.3.1 A run chart is a plot of results in chronological order
that can be used to screen data for unusual patterns. Preferably,
pretreated results are plotted. Use a run chart to screen data for
unusual patterns such as continuous trending in either direc-
tion, unusual clustering, and cycles. Several non-random pat-
terns are described in control chart literature. When control
parameters have been added to a run chart, it becomes a control
chart of individual values (I chart).
A1.3.2 Plot results on the chart. Plot the first result at the
left, and plot each subsequent point one increment to the right
of its predecessor. The points may be connected in sequence to
facilitate interpretation of the run chart.
A1.3.3 Allow sufficient space in the x-axis direction to
accommodate as many results as should be obtained from a FIG. A1.2 Run Chart for Multiple Results from a Single Check
Standard
consistent batch of material. Allow enough space in the y-axis
direction to accommodate the expected minimum and maxi-
mum of the data.
A1.3.4 Example of a Run Chart for QC Results—The first
15 results from Column 2 of Table A1.1 are plotted in sequence
as they are collected as shown in Fig. A1.1. The data would be
examined for unusual patterns.
A1.3.5 Example of a Run Chart for Multiple Results from a
Single Check Standard—The first 15 preprocessed results
(differences) from Column 4 of Table A1.2 are plotted in
sequence as they are collected as shown in Fig. A1.2. The data
would be examined for unusual patterns.
A1.3.6 Example of a Run Chart for Results from Multiple
Check Standards—The first 15 preprocessed results (differ-
ences scaled by si) from Table A1.3 are plotted in sequence as
they are collected as shown in Fig. A1.3. The data would be FIG. A1.3 Run Chart for Results from Multiple Check Standards
examined for unusual patterns.
(2) Select the appropriate column from Fig. A1.4, based on
A1.4 Normality Checks the number of observations (n).
A1.4.1 A normal probability plot (a special case of a q-q (3) Plot each observation in the sorted column (y-value)
plot) is used to test the assumption that the observations are against its corresponding value from Fig. A1.4 (z-value).
normally distributed. Since the control chart and limits pre- A1.4.1.2 Visually inspect the plot for an approximately
scribed in this practice are based on the assumption that the linear relationship. If the results are normally distributed, the
data behavior is adequately modeled by the normal distribu- plot should be approximately linear. Major deviations from
tion, it is recommended that a test of this normality assumption linearity are an indication of nonnormal distributions of the
be conducted. differences.
A1.4.1.1 To construct a normal probability plot:
NOTE A1.1—The assessment methodology of the normal probability
(1) Create a column of the observations sorted in ascending plot advocated in this practice is strictly visual due to its simplicity. For
order. statistically more rigorous assessment techniques, users are advised to
consult a statistician.
A1.4.2 Anderson-Darling Statistic—The Anderson-Darling
statistic is used to test for normality. The test involves the
following steps:
A1.4.2.1 Order the non-outlying results such that x1 # x2 #
. . . . xn
A1.4.2.2 Obtain standardized variate from the xi values as
follows:
wi 5 ~xi – x̄!/s (A1.4)
for (i= 1 . . . n), where s is sample standard deviation of the
results, and x̄ is the average of the results.
A1.4.2.3 Convert the wi values to standard normal cumula-
tive probabilities pi values using the cumulative probability
FIG. A1.1 Example of a Run Chart for QC Results table for the standardized normal variate z (see Fig. A1.5):

9
 D 6299 – 02

FIG. A1.4 z–Values

pi 5 Probability ~z , wi! (A1.5) the calculation of the Anderson-Darling statistic are shown in
A1.4.2.4 Compute A as: 2 Table A1.4, as is the individual terms in the summation for A2.
n
The value for A2 is 0.415, and the value for A2* is 0.440. Since
( ~2i – 1! @ln~pi! 1 ln~1 – pn 1 1 – i!#
i51
this value is less than 0.752, the hypothesis of normality is
A2 5 – –n (A1.6) accepted at the 95 % confidence level.
n
A1.4.4 Example of Normal Probability Plot for Multiple
A1.4.2.5 Compute A2* as: Results from a Single Check Standard—The first 15 prepro-
S 0.75 2.25
A2 * 5 A 2 1 1 n 1 2
n D (A1.7)
cessed results (Table A1.2, Column 4) are sorted in ascending
order and paired with the corresponding z-values from Fig.
A1.4.2.6 If the computed value of A2* exceeds 0.752, then A1.4. The paired results (Table A1.5) are plotted as x,y points
the hypothesis of normality is rejected for a 5 % level test. (Fig. A1.7). A line can be added to the plot to facilitate
A1.4.3 Example of Normal Probability Plot for QC examination of the data for deviations from linearity.
Results—Once 15 results have been obtained (Table A1.1), A1.4.4.1 For this example, the wi, and pi values used in the
they are sorted in ascending order and paired with the calculation of the Anderson-Darling statistic are shown in
corresponding z-values from Fig. A1.4. The paired results (see Table A1.6, as are the individual terms in the summation for A2.
Table A1.4) are plotted as (x,y) points (see Fig. A1.6). A line The value for A2 is 0.415, and the value for A2* is 0.440. Since
can be added to the plot to facilitate examination of the data for this value is less than 0.752, the hypothesis of normality is
deviations from linearity. accepted at the 95 % confidence level.
A1.4.3.1 For the above example, the wi and pi values used in A1.4.5 Example of Normal Probability Plot for Results from

10
 D 6299 – 02

FIG. A1.4 z–Values (continued)

Multiple Check Standards—The first 15 preprocessed results data for deviations from linearity.
(Table A1.3, Column 6) are sorted in ascending order and
paired with the corresponding z-values from Fig. A1.4. The A1.5 The Control Chart
paired results (Table A1.7) are plotted as x,y points (Fig. A1.8). A1.5.1 I Chart—The I chart is a run chart to which control
A line can be added to the plot to facilitate examination of the limits and center line have been added. To establish placement

11
 D 6299 – 02

NOTE—Probability (z < w i), where wi is the sum of the number in the left column and top row.
FIG. A1.5 pi Values

positions of the control limits for the I chart, an estimate of the

Ī 5
(i51
n
Ii
(A1.8)
process variability will need to be obtained from the data. n
While there are several statistical techniques that can be used
A1.5.1.2 Upper and lower control limits are added, also,
for this purpose, this practice advocates use of an MR of two
chart for its simplicity and robustness to outliers. Produce an I computed from the MR of two:
chart only after a minimum of 15 preprocessed results have
MR 5
(i51
n–1
?Ii11 – Ii?
(A1.9)
been obtained from the measurement system, and the data have n–1
been screened (see 8.3.1 and 8.3.2) and tested for normality
(see A1.4). UCLI 5 Ī 1 2.66 MR (A1.10)
A1.5.1.1 A horizontal center line is added at the level of the
mean of all the results, Ī:

12
 D 6299 – 02

FIG. A1.5 pi Values (continued)

¯
LCLI 5 Ī – 2.66 MR (A1.11) A1.5.2 MR Chart—A MR of two chart is obtained by
A1.5.1.3 Individual values that are outside the upper or plotting the sequential range of two values given by:
lower control limits are indications of an unstable system, and MRi 5 ?Ii – Ii–1? (A1.12)
efforts should be made to determine the cause. Optionally, any and connecting each point.
one of the following occurrences should be considered as A1.5.2.1 The upper control limit for the MR chart is given
potential signs of instability: by:
(1) Two out of three consecutive results on the I chart that
UCLMR 5 3.27 MR (A1.13)
are more than 1.77 MR distant from the center line in the same
direction; A1.5.2.2 There is no lower control limit for an MR chart.
(2) Five consecutive results on the I chart that are more than A1.5.3 EWMA Overlay—A EWMA overlay is a trend line
0.89 MR distant from the center line in the same direction; constructed from EWMA values calculated using the I-values.
(3) Eight or more consecutive points in the I chart that fall on The EWMA trend line is typically overlaid on the I chart to
the same side of the center line. enhance its sensitivity in detecting mean shifts that are small

13
 D 6299 – 02
TABLE A1.4 Example Data for a Normal Probability Plot for QC
Results
Original
ith Term in Eq
Sequence z-value Sorted Result wi pi
A1.6
No., I
11 -1.83 55.0 -1.47 0.07 -5.91
14 -1.28 55.2 -1.07 0.14 -14.35
1 -0.97 55.3 -0.86 0.19 -18.70
7 -0.73 55.3 -0.86 0.19 -21.94
8 -0.52 55.4 -0.66 0.25 -25.77
6 -0.34 55.5 -0.46 0.32 -21.44
12 -0.17 55.5 -0.46 0.32 -25.34
13 0.00 55.5 -0.46 0.32 -22.80
2 0.17 55.8 0.15 0.56 -16.52
5 0.34 55.8 0.15 0.56 -18.46
10 0.52 56.1 0.76 0.78 -11.50
4 0.73 56.1 0.76 0.78 -10.80
3 0.97 56.3 1.16 0.88 -8.65
15 1.28 56.5 1.57 0.94 -5.79 FIG. A1.7 Example of a Normal Probability Plot for Multiple
9 1.83 56.6 1.77 0.96 -3.25 Results from a Single Check Standard

TABLE A1.6 Example Data for a Normal Probability Plot for

Results from Multiple Check Standards
Original
Sorted ith Term in
Sort No. Sequence z-value wi pi
Result Eq A1.6
No.
1 11 -1.83 -1.35 -3.12 0.00 -10.41
2 14 -1.28 -0.83 -1.94 0.03 -15.11
3 1 -0.97 -0.56 -1.32 0.09 -18.58
4 7 -0.73 -0.53 -1.25 0.11 -24.98
5 8 -0.52 -0.35 -0.84 0.20 -25.59
6 6 -0.34 0.01 -0.02 0.49 -19.68
7 12 -0.17 0.09 0.16 0.56 -19.93
8 13 0 0.09 0.16 0.56 -21.05
9 2 0.17 0.15 0.30 0.62 -22.33
10 5 0.34 0.2 0.41 0.66 -20.75
11 10 0.52 0.26 0.55 0.71 -11.91
12 4 0.73 0.29 0.62 0.73 -9.73
13 3 0.97 0.3 0.64 0.74 -9.99
14 15 1.28 0.32 0.69 0.75 -8.34
15 9 1.83 0.82 1.83 0.97 -1.02
FIG. A1.6 Example of a Normal Probability Plot for QC Results

TABLE A1.5 Example Data for a Normal Probability Plot for overlaid on the I chart and connected. Use the following
Multiple Results from a Single Check Standard recursion equation:
Original
Sort No. Sequence
Sorted
z-value wi pi
ith Term in EWMA1 5 I1 (A1.14)
Result Eq A1.6
No.
1 11 -0.88 -1.83 -1.47 0.07 -5.91 EWMAi 5 ~1 – l!EWMAi–1 1 lIi (A1.15)
2 14 -0.68 -1.28 -1.07 0.14 -14.35
3 1 -0.58 -0.97 -0.86 0.19 -18.70
where l is the exponential weighting factor. For application
4 7 -0.58 -0.73 -0.86 0.19 -21.94 of this practice, a l value of 0.4 is recommended.
5 8 -0.48 -0.52 -0.66 0.25 -25.77
6 6 -0.38 -0.34 -0.46 0.32 -21.44 NOTE A1.2—For the EWMA trend, a l value of 0.4 closely emulates the
7 12 -0.38 -0.17 -0.46 0.32 -25.34 run rule effects of conventional control charts, while a value of 0.2 has
8 13 -0.38 0 -0.46 0.32 -22.80 optimal prediction properties for the next expected value. In addition,
9 2 -0.08 0.17 0.15 0.56 -16.52 these l values also conveniently places the control limits (3-sigma) for the
10 5 -0.08 0.34 0.15 0.56 -18.46
11 10 0.22 0.52 0.76 0.78 -11.50
EWMA trend at the 1 (for l=0.2) to 1.5-sigma (for l=0.4) values for I
12 4 0.22 0.73 0.76 0.78 -10.80 chart.
13 3 0.42 0.97 1.16 0.88 -8.65
14 15 0.62 1.28 1.57 0.94 -5.79
A1.5.3.2 The control limits for the EWMA chart are calcu-
15 9 0.72 1.83 1.77 0.96 -3.25 lated using a weight (l) as follows:

UCLl 5 Ī 1 2.66 MR Œ l
2–l (A1.16)

relative to the measurement system precision. Each EWMA

value is a weighted average of the current result and previous
results, with the weights decreasing exponentially with the age
LCLl 5 Ī – 2.66 MR Œ l
2–l (A1.17)

of the reading. A1.5.4 Examples of Control Charts for QC and Check

A1.5.3.1 A sequence of values, EWMA i, are calculated, and Standard Results:

14
 D 6299 – 02
TABLE A1.7 Example Data for I Chart and EWMA Overlay for QC
Results
Sequence Number,
QC Result (Yi=Ii) Moving Range MRi EWMAi
I
1 55.3 55.3
2 55.8 0.5 55.50
3 56.3 0.5 55.82
4 56.1 0.2 55.93
5 55.8 0.3 55.88
6 55.5 0.3 55.73
7 55.3 0.2 55.56
8 55.4 0.1 55.49
9 56.6 1.2 55.94
10 56.1 0.5 56.00
11 55 1.1 55.60
12 55.5 0.5 55.56
13 55.5 0.0 55.54
14 55.2 0.3 55.40
15 56.5 1.3 55.84
FIG. A1.9 Example of a MR Chart for QC Results
Average 55.73 0.500

16 55.7 0.8 55.78

17 55.6 0.1 55.71
18 55.2 0.4 55.51
19 55.7 0.5 55.58
20 56.1 0.4 55.79
21 56.3 0.2 55.99
22 55.2 1.1 55.68
23 55.4 0.2 55.57
24 55.4 0.0 55.50
25 55.6 0.2 55.54

FIG. A1.10 Example of an I-Chart with EWMA Overlay for QC

Results

plotted in sequence. After 15 results are obtained (Table A1.2),

the MR value is calculated and added to the plot. A UCLMR is
added to produce the MR chart (see Fig. A1.11).
A1.5.4.4 Example of I Chart and EWMA Overlay for
FIG. A1.8 Example of a Normal Probability Plot for Results from Multiple Results from a Single Check Standard—The average
Multiple Check Standards of the first 15 QC results (see Table A1.2, Column 4) is

A1.5.4.1 Example of a MR Chart for QC Results—MRi

values for the data from Table A1.1 are calculated and plotted
in sequence. After 15 results are obtained, the MR = 0.500
value is calculated and added to the plot. Computations are
shown in Table A1.7. A UCLMR=1.64 is added to produce the
MR chart (Fig. A1.9).
A1.5.4.2 Example of I Chart and EWMA Overlay for QC
Results—The average of the first 15 QC results (Table A1.7,
Column 2) is calculated and plotted on the run chart as Ī
=55.73. The upper and lower control limits are calculated from
Eq A1.10 and Eq A1.11 as 54.40 and 57.06 and added to the
run chart to produce the I chart (Fig. A1.10). EWMA values
(Table A1.7, Column 4) and EWMA control limits, 55.06 and
56.39, are overlaid on the I chart. Additional results and
calculated EWMA values are added as they are determined.
A1.5.4.3 Example of a MR Chart for Multiple Results from FIG. A1.11 Example of a MR Chart for Multiple Results from a
a Single Check Standard—MRi values are calculated and Single Check Standard

15
 D 6299 – 02
calculated and plotted on the run chart as Ī. The upper and TABLE A1.8 Example Data for I Chart and EWMA Overlay for
lower control limits are calculated from Eq A1.6 and Eq A1.7 Multiple Results from a Single Check Standard
and added to the run chart to produce the I chart. EWMA values Check Standard
Sequence Number Moving Range, MRi EWMAi
Result (Ii)
and EWMA control limits may be overlaid on the I chart (Fig.
A1.12). Additional results and calculated EWMA values are 1 -0.58 -0.58
2 -0.08 0.5 -0.38
added as they are determined. The MR values for this example 3 0.42 0.5 -0.06
are shown in Table A1.8, Column 3.) 4 0.22 0.2 0.05
5 -0.08 0.3 -0.00
A1.5.4.5 Example of a MR Chart for Results from Multiple 6 -0.38 0.3 -0.15
Check Standards—MRi values are calculated and plotted in 7 -0.58 0.2 -0.32
sequence. After 15 results are obtained (Table A1.3, Column 6, 8 -0.48 0.1 -0.39
9 0.72 1.2 0.06
displayed again in Table A1.9), the MR value is calculated and 10 0.22 0.5 0.12
added to the plot. A UCLMR is added to produce the MR chart 11 -0.88 1.1 -0.28
(see Fig. A1.13). 12 -0.38 0.5 -0.32
13 -0.38 0.0 -0.34
A1.5.4.6 Example of I Chart and EWMA Overlay for 14 -0.68 0.3 -0.48
Results from Multiple Check Standards—The average of the 15 0.62 1.3 -0.04
first 15 QC results (see Table A1.3, Column 6) is calculated and
Average -0.153 0.500
plotted on the run chart as Ī. The upper and lower control limits
are calculated from Eq A1.10 and Eq A1.11 and added to the 16 -0.18 0.8 -0.10
run chart to produce the I chart. EWMA values and EWMA 17 -0.28 0.1 -0.17
18 -0.68 0.4 -0.37
control limits may be overlaid on the I chart (Fig. A1.14). 19 -0.18 0.5 -0.30
Additional results and calculated EWMA values are added as 20 0.22 0.4 -0.09
they are determined. 21 0.42 0.2 0.11
22 -0.68 1.1 -0.20
A1.6 t Test 23 -0.48 0.2 -0.31
24 -0.48 0.0 -0.38
A1.6.1 A two sided t test is used to check if a sample of 25 -0.28 0.2 -0.34
values comes from a population with a mean different from an
hypothesized value, µ0. In this practice, a t test may be
performed on pretreated check standard test results to check for TABLE A1.9 Example Data for a MR Chart for Results from
Multiple Check Standards
bias relative to the ARVs. Since during pretreatment, accepted
reference value(s) have been subtracted from the raw results, Result Sequence Preprocessed
Moving Range, MRi EWMAi
Number, i Result, Ii
the hypothesized mean value is zero.
1 -0.35 -0.35
A1.6.1.1 For the purpose of performing the t test, two 2 0.82 1.17 0.12
methods for calculating the t value are presented: 3 0.09 0.73 0.11
(1) By the root-mean square method, the standard deviation 4 -1.35 1.44 -0.48
5 0.32 1.67 -0.16
of the pretreated results is calculated as:

Œ
6 -0.83 1.15 -0.43
n 7 0.30 1.13 -0.14
( ~Ii – Ī!2
i51
8
9
-0.53
0.15
0.83
0.68
-0.29
-0.12
SI 5 n–1 (A1.18) 10 0.09 0.06 -0.03
11 0.26 0.17 0.08
(2)The t value is calculated as: 12 -0.56 0.82 -0.17
13 0.20 0.76 -0.02
t 5 =n?Ī – µ0? /S I (A1.19) 14
15
0.01
0.29
0.19
0.28
-0.01
0.11

Average -0.073 0.791

16 0.59 0.3 0.30

17 -1.19 1.78 -0.29
18 -0.13 1.06 -0.23
19 -0.41 0.28 -0.30
20 -0.73 0.32 -0.47
21 0.14 0.87 -0.23
22 -0.38 0.52 -0.29
23 -0.7 0.32 -0.45
24 0.17 0.87 -0.20

where µ0 is the hypothesized mean, which is zero (see

A1.6.1).
(3) Alternatively, by the MR approach, compute the alternate
t value as:
FIG. A1.12 Example of an I-Chart with EWMA Overlay for Multiple
Results from a Single Check Standard tMR 5 =n?Ī – µ0? / ~MR/1.128! (A1.20)

16
 D 6299 – 02

FIG. A1.13 Example of a MR Chart for Results from Multiple

Check Standards

FIG. A1.14 Example of an I-Chart with EWMA Overlay for Results

from Multiple Check Standards

where µ0 is the hypothesized mean, which is zero (see freedom. If tMR from Eq A1.20 is used, the appropriate degrees
A1.6.1). of freedom are (n–1)/2.
A1.6.1.2 Compare the computed t value from Eq A1.19
with the critical t values in Table A1.10 for (n–1) degrees of

TABLE A1.10 95th Percentile of Student’s ?t? Distribution

Degrees
t
of Freedom
1 12.7062
2 4.3027
3 3.1824
4 2.7764
5 2.5706
6 2.4469
7 2.3646
8 2.3060
9 2.2622
10 2.2281
11 2.2010
12 2.1788
13 2.1604
14 2.1448
15 2.1314
16 2.1199
17 2.1098
18 2.1009
19 2.0930
20 2.0860
21 2.0796

17
 D 6299 – 02

TABLE A1.10 Continued

Degrees
t
of Freedom
22 2.0739
23 2.0687
24 2.0639
25 2.0595
26 2.0555
27 2.0518
28 2.0484
29 2.0452
30 2.0423
31 2.0395
32 2.0369
33 2.0345
34 2.0322
35 2.0301
36 2.0281
37 2.0262
38 2.0244
39 2.0227
40 2.0211
41 2.0195
42 2.0181
43 2.0167
44 2.0154
45 2.0141
46 2.0129
47 2.0117
48 2.0106
49 2.0096
50 2.0086
55 2.0040
60 2.0003
65 1.9971
70 1.9944
75 1.9921
80 1.99006
85 1.98827
90 1.98667
95 1.98525
100 1.98397
105 1.98282
110 1.98177
115 1.98081
120 1.97993
125 1.97912
130 1.97838
135 1.97769
140 1.97705
145 1.97646
150 1.97591
155 1.97539
160 1.97490
165 1.97445
170 1.97402
175 1.97361
180 1.97323
185 1.97287
190 1.97253
195 1.97220
200 1.97190

A1.6.1.3 If the absolute value of the calculated t (or tMR) bias in the measurement system.
value is less than or equal to the critical t value, then µ0 is A1.6.2 Example of t Test Applied to Multiple Results from a
statistically indistinguishable from the mean of the distribution. Single Check Standard—For the first 15 preprocessed results in
For the case of check standard testing, this would indicate that Column 4 of Table A1.2, Ī is –0.153. Since the results being
there is no statistically identifiable bias. analyzed are the difference relative to the ARV, µ0 is zero. The
A1.6.1.4 If the absolute value of t is greater than the critical standard deviation of the first 15 preprocessed results is 0.493,
t value, then µ0 is statistically distinguishable from the mean of and the t value is 1.2034. The t value is less than the critical
the distribution, with 95 % confidence. For the case of check value for 14 degrees of freedom (t14= 2.1448), so the average
standard testing, this would indicate a statistically identifiable difference between the check standard results and the accepted

18
 D 6299 – 02
reference value is statistically indistinguishable from zero. A1.7.4 Example—The site precision calculated from R8=
A1.6.3 Example of t Test Applied to Results from Multiple 2.46 MR for the first 15 QC results in Table A1.1 is 1.23. The
Check Standards—For the first 15 preprocessed results in published reproducibility for the measurement method at the
Column 6 of Table A1.3, Ī is –0.0719. Since the results being 58.88 level is 1.05. x2 is therefore 14·1.232/2·1.052 = 11.57.
analyzed are the difference relative to the ARV, µ0 is zero. The This value is less than the critical x2 value of 14.1 for 7 degrees
standard deviation of the first 15 preprocessed results is 0.550, of freedom, so the site precision is not statistically greater than
and the t value is 0.506. The t value is less than the critical the published reproducibility of the method.
value for 14 degrees of freedom (t14 = 2.1448), so the average A1.8 Approximate F Test
difference between the check standard results and the accepted
reference value is statistically indistinguishable from zero. A1.8.1 In this practice, an approximate F test is used to
compare the variation exhibited by a measurement system over
A1.7 Approximate Chi-Square Test two different time periods. It can also be used to compare the
site precision estimated from a series of results from one QC
A1.7.1 The chi-square (x2) test is used to compare the
sample with that estimated using a different QC sample (see
estimated site precision to a published reproducibility value, as
8.6.1).
instructed in 9.1.2.
A1.8.2 Compute the F value as:
A1.7.2 Compute the chi-square statistic:
2 ~n – 1!R8 2
F 5 MR 12 / MR 2
2
(A1.22)
x 5 (A1.21) where MR1 is the larger of the two average moving ranges,
2R2
and MR2 is the smaller.
where R8 is the estimated site precision (R8=2.46 MR ) and A1.8.3 Compare the computed F value to the critical F
R is the published reproducibility of the method. value read from Table A1.12, with (n1-1)/2 degrees of freedom
A1.7.3 Compare the computed x2 value to the critical x2 for the numerator and (n2-1)/2 degrees of freedom for the
value in Table A1.11, with (n–1)/2 degrees of freedom. If n is denominator.
even, interpolate. A1.8.3.1 If the computed F value exceeds the tabled value,
A1.7.3.1 If the computed x2 value exceeds the tabled value, then the two precisions are statistically distinguishable. We can
then the site precision exceeds the published reproducibility of be 95 % confident that the process that gave rise to the moving
the method, with 95 % confidence. range MR1 is less precise (has larger site precision) than the
A1.7.3.2 If the computed x2 value is less than or equal to the process that produced MR2 .
tabled value, then the site precision is either less than or A1.8.3.2 If the computed F value is smaller than the tabled
statistically indistinguishable from the published reproducibil- value, then the precisions of the two samplings of the mea-
ity of the test method. surement process are statistically indistinguishable.
NOTE A1.3—Although the approximate F-test is conducted at the 95 %
TABLE A1.11 95th Percentiles of the Chi Square Distribution probability level, the critical F values against which the calculated F is
Degrees compared come from the 97.5 percentiles of the F-statistic. If the ratio
X
Freedom MRa2 / MRb2 is calculated without requiring that the larger variance is in
7 14.1 the numerator, the ratio would have to be compared against both the lower
8 15.5 2.5 percentile point and the upper 97.5 percentile point of the
9 16.9 F-distribution to determine if the two variances were statistically distin-
10 18.3 guishable. Because of the nature of the F-distribution, comparing MRa2 /
11 19.7
12 21.0
MRb2 to the 2.5 percentile point when MRa2 / MRb2 is equivalent to
13 22.4 comparing MRb2 / MRa2 to the 97.5 percentile point. Requiring that larger
14 23.7 variance is always in the numerator allows the“ two-tailed” test to be
15 25.0 accomplished in one step. If the variance of the two populations were
16 26.3 equal, then there would be only a 2.5 % chance that MR12 > MR22 by more
17 27.6
than the tabulated amount, and a 2.5 % chance that MR22 > MR12 by more
18 28.9
19 30.1 than the tabulated amount with degrees of freedom reversed.
20 31.4 A1.8.4 If two precision estimates are statistically indistin-
21 32.7
22 33.9
guishable, they may be pooled into a single estimate. For
23 35.2 example, if MR1 was obtained from measurements on a single
24 36.4 lot of QC sample material, while MR2 was obtained from
25 37.7
26 38.9
measurements on a different lot of material, and, if they are not
27 40.1 statistically distinguishable, they may be pooled. The appro-
28 41.3 priate pooled precision estimate is
30 43.8
35 49.8 ~n1 – 1!MR1 1 ~n2 – 1! MR2
40 55.8 MRpooled 5 n1 1 n2 – 2 (A1.23)
45 61.7
50 67.5 A1.8.5 Example—Table A1.13 contains QC results for a
60 79.1 second QC sample measured by the same measurement system
70 90.5
80 101.9 used to generate the results in Table A1.1. The MR value for the
25 results from the original QC sample (Table A1.1) was 0.454.

19
 D 6299 – 02
TABLE A1.12 97.5 Percentiles of the F Statistic
Denominator, Numerator
degrees of
7 8 9 10 12 14 16 18 20 25 30 40 50 100
freedom
7 4.99 4.90 4.82 4.76 4.67 4.60 4.54 4.50 4.47 4.40 4.36 4.31 4.28 4.21
8 4.53 4.43 4.36 4.30 4.20 4.13 4.08 4.03 4.00 3.94 3.89 3.84 3.81 3.74
9 4.20 4.10 4.03 3.96 3.87 3.80 3.74 3.70 3.67 3.60 3.56 3.51 3.47 3.40
10 3.95 3.85 3.78 3.72 3.62 3.55 3.50 3.45 3.42 3.35 3.31 3.26 3.22 3.15
11 3.76 3.66 3.59 3.53 3.43 3.36 3.30 3.26 3.23 3.16 3.12 3.06 3.03 2.96
12 3.61 3.51 3.44 3.37 3.28 3.21 3.15 3.11 3.07 3.01 2.96 2.91 2.87 2.80
13 3.48 3.39 3.31 3.25 3.15 3.08 3.03 2.98 2.95 2.88 2.84 2.78 2.74 2.67
14 3.38 3.29 3.21 3.15 3.05 2.98 2.92 2.88 2.84 2.78 2.73 2.67 2.64 2.56
15 3.29 3.20 3.12 3.06 2.96 2.89 2.84 2.79 2.76 2.69 2.64 2.59 2.55 2.47
16 3.22 3.12 3.05 2.99 2.89 2.82 2.76 2.72 2.68 2.61 2.57 2.51 2.47 2.40
17 3.16 3.06 2.98 2.92 2.82 2.75 2.70 2.65 2.62 2.55 2.50 2.44 2.41 2.33
18 3.10 3.01 2.93 2.87 2.77 2.70 2.64 2.60 2.56 2.49 2.44 2.38 2.35 2.27
19 3.05 2.96 2.88 2.82 2.72 2.65 2.59 2.55 2.51 2.44 2.39 2.33 2.30 2.22
20 3.01 2.91 2.84 2.77 2.68 2.60 2.55 2.50 2.46 2.40 2.35 2.29 2.25 2.17
25 2.85 2.75 2.68 2.61 2.51 2.44 2.38 2.34 2.30 2.23 2.18 2.12 2.08 2.00
30 2.75 2.65 2.57 2.51 2.41 2.34 2.28 2.23 2.20 2.12 2.07 2.01 1.97 1.88
35 2.68 2.58 2.50 2.44 2.34 2.27 2.21 2.16 2.12 2.05 2.00 1.93 1.89 1.80
40 2.62 2.53 2.45 2.39 2.29 2.21 2.15 2.11 2.07 1.99 1.94 1.88 1.83 1.74
45 2.58 2.49 2.41 2.35 2.25 2.17 2.11 2.07 2.03 1.95 1.90 1.83 1.79 1.69
50 2.55 2.46 2.38 2.32 2.22 2.14 2.08 2.03 1.99 1.92 1.87 1.80 1.75 1.66
60 2.51 2.41 2.33 2.27 2.17 2.09 2.03 1.98 1.94 1.87 1.82 1.74 1.70 1.60
70 2.47 2.38 2.30 2.24 2.14 2.06 2.00 1.95 1.91 1.83 1.78 1.71 1.66 1.56
80 2.45 2.35 2.28 2.21 2.11 2.03 1.97 1.92 1.88 1.81 1.75 1.68 1.63 1.53
90 2.43 2.34 2.26 2.19 2.09 2.02 1.95 1.91 1.86 1.79 1.73 1.66 1.61 1.50
100 2.42 2.32 2.24 2.18 2.08 2.00 1.94 1.89 1.85 1.77 1.71 1.64 1.59 1.48

TABLE A1.13 Example of QC Results for a Second QC Sample Plot the result from the old material on its I chart, MR chart,
Measured by the Same Measurement System EWMA chart, or Q chart, or a combination of these. If no
Sequence special-cause signals are noted, then the result for the new
QC Result MR Cn LCL UCL
Number
material is considered to be valid.
1 54.2 A1.9.3 Plot the result from the new material as the first point
2 56.1 1.9 55.15 54.21 56.09
3 55.2 0.9 55.17 54.08 56.25 on the Q chart.
4 54.1 1.1 54.90 53.75 56.05
5 53.7 0.4 54.66 53.47 55.85 NOTE A1.4—The Q chart is essentially a control chart of transformed
6 54 0.3 54.55 53.34 55.76 statistics calculated from the conventional statistics normally plotted on
7 54.3 0.3 54.51 53.28 55.75 control charts (for example, mean, range). This transformed statistics
8 54.8 0.5 54.55 53.31 55.79 retains the information from the conventional statistics, but has the
9 53.9 0.9 54.48 53.22 55.73 advantage of permitting plotting of all points on one standardized control
10 53.2 0.7 54.35 53.09 55.61
chart.
11 52.5 0.7 54.18 52.91 55.45
12 52.8 0.3 54.07 52.79 55.34 A1.9.3.1 Center this value on the y-axis of the new chart.
13 54.3 1.5 54.08 52.81 55.36
14 52.7 1.6 53.99 52.70 55.27
Scale the y-axis to allow room for the initial result plus and
15 53.4 0.7 53.95 52.66 55.23 minus five historical standard deviations, where the standard
16 53.1 0.3 53.89 52.61 55.18 deviations are appropriate to the level of the first result.
17 54 0.9 53.90 52.61 55.19
18 53.2 0.8 53.86 52.57 55.15
A1.9.3.2 No center line, nor upper or lower control limits,
19 52.8 0.4 53.81 52.51 55.10 are plotted at this time.
20 53.2 0.4 53.78 52.48 55.07 A1.9.4 Subsequent QC sample testing may be done only on
21 53.1 0.1 53.74 52.44 55.04
22 53.3 0.2 53.72 52.42 55.02
the new material.
23 52.8 0.5 53.68 52.38 54.98 A1.9.5 Plot subsequent QC results as points on the new Q
chart. Do not connect the points.
A1.9.6 As each point (the nth point) is plotted, compute and
The MR value for the 23 results for the new QC sample is plot the center value and the upper and lower control limits
0.700. The F value is 2.38, which is less than the critical value applicable for this result.
of 3.33 for 11 and 12 degrees of freedom in the numerator and
denominator, respectively. The precision of the measurements
A1.9.6.1 Center value, Cn 5 (i51 n
/
Ii n , where the sum
includes the latest result, In. Optionally plot and connect the
for the two QC batches is statistically indistinguishable.
sequence of points {Cn} with a broken line. (Alternatively,
A1.9 Q-Procedure replace any previous center line with a new line at the latest
A1.9.1 Collect and prepare a new batch of QC material value of Cn.)
while the current QC material supply remaining can support at A1.9.6.2 Upper control limit, UCLn= Cn +
least two additional analyses. 3s = ~n–1! / n , where s is the historical standard deviation
A1.9.2 Concurrently test the first sample of the new material appropriate for test level Cn. For example, if the standard
with a routine analysis of the soon-to-be-depleted QC material. deviation is unchanged from the exhausted QC sample, then s

20
 D 6299 – 02
= MR /1.128. Connect the sequence of points {UCLi} with a EWMAi 5 ~1 – l!EWMAi21 1 lIi (A1.25)
broken line. (Alternatively, replace any previous upper control where l is the exponential weighting factor, typically set to
limit lines with a new line at the latest UCLn. 0.4.
A1.9.6.3 Lower control limit, LCLn= Cn – 3 s = ~n–1! / n A1.9.9.2 The upper control limit for the EWMA chart is
. Connect the sequence of points {LCLi} with a broken line.
(Alternatively, replace any previous lower control limit lines UCLEWMA 5 Cn 1 3s ŒS D S D
l 1–l
2–l 1 2 2–l ~1 – l!
2~n–1! 1
–n
with a new line at the latest LCLn.) (A1.26)
A1.9.7 Individual values, current or earlier, which are out- A1.9.9.3 The lower control limit for the EWMA chart is
side the current upper or lower control limits, are indications of
an unstable system, and efforts should be made to determine
the cause. Optional run rules (corresponding to A1.5.1.3(1) to
LCLEWMA 5 Cn – 3s ŒS D S D
l 1–l
2–l 12 2–l ~1 – l!
2~n–1! 1
–n
(A1.27)
(3)) may also be applied to sequences of points using the
current UCL and LCL, as early indicators of instability: A1.9.10 Example—It is assumed that the collection of the
A1.9.7.1 Two consecutive results on the Q chart that are QC results in Table A1.13 was started when there was sufficient
more than 2 s = ~n–1! / n distant from the current expected
quantity of QC batch 1 (Table A1.1) for two analyses. The
individual values are plotted as they are collected, (squares in
value, Cn, in the same direction;
Fig. A1.15), and the Cn and UCLn and LCLn values are
A1.9.7.2 Five consecutive results on the Q chart that are
calculated and added for each new result. Recall that MR from
more than s = ~n–1! / n distant from the current expected the first 15 measurements on batch 1 was 0.500. The new
value in the same direction. control limits (Table A1.13, Columns 5 and 6) are compared to
A1.9.7.3 Eight consecutive results on the Q chart that are on the current and previous results. Note that, for this example, the
the same side of the current expected value. second result is considered “out of control” when UCL is
A1.9.8 Continue or replace the MR chart, as appropriate. calculated. The “out-of-control” character of this result is
A1.9.8.1 If the standard deviation for the new QC material confirmed as UCL is updated with additional data. The Q chart
is the same as for the old material, continue the old MR chart clearly shows that the results for the new QC sample trend
beginning with MR2, that is, the second result from the new downward with time.
material.
A1.9.8.2 If the standard deviation appropriate to the level of
the new material is different from the old, begin a new MR
chart, starting with MR2. The upper control limit for the new
chart should be placed at 3.69s.
A1.9.8.3 After 15 results have been obtained with the new
material, use a chi-square (see A1.7) or F test (see A1.8) to
check that s is appropriate for the new material.
A1.9.9 EWMA Overlay on a Q Chart—An EWMA chart
may be overlaid on a Q chart, although it will not be
meaningful until n > 5.
A1.9.9.1 The sequence of EWMA values, EWMAi, are
calculated, and overlaid on the I chart and connected. Use the
following recursion:
EWMA1 5 I1 (A1.24)
FIG. A1.15 Example of a Q-Chart for a New QC Sample

21
 D 6299 – 02

REFERENCES

(1) “Total Quality Management,” ASTM STP 1209, ASTM. (5) Hunter, J. S., “A One-Point Plot Equivalent to the Shewhart Chart with
(2) Manual on Presentation of Data and Control Chart Analysis, Manual Western Electric Rules,” Quality Engineering, Vol 2, No. 1, 1989-
7, ASTM STP 15, ASTM. 1990, pp. 13-19.
(3) “Glossary and Tables for Statistical Quality Control,” American (6) Quesenberry, C. P., “SPC Q-Charts for Start-up Processes and Short or
Society for Quality Control. Long Runs,” Journal of Quality Technology, Vol 23, No. 3, July 1991,
(4) Hunter, J. S., “The Exponentially Weighted Moving Average,” Journal
pp. 213-224.
of Quality Technology, Vol 18, No. 4, October 1986, pp. 203-210.

ASTM International takes no position respecting the validity of any patent rights asserted in connection with any item mentioned
in this standard. Users of this standard are expressly advised that determination of the validity of any such patent rights, and the risk
of infringement of such rights, are entirely their own responsibility.

This standard is subject to revision at any time by the responsible technical committee and must be reviewed every five years and
if not revised, either reapproved or withdrawn. Your comments are invited either for revision of this standard or for additional standards
and should be addressed to ASTM International Headquarters. Your comments will receive careful consideration at a meeting of the
responsible technical committee, which you may attend. If you feel that your comments have not received a fair hearing you should
make your views known to the ASTM Committee on Standards, at the address shown below.

This standard is copyrighted by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959,
United States. Individual reprints (single or multiple copies) of this standard may be obtained by contacting ASTM at the above
address or at 610-832-9585 (phone), 610-832-9555 (fax), or [email protected] (e-mail); or through the ASTM website
(www.astm.org).

22