ASTM D4467-94
(Practice)Standard Practice for Interlaboratory Testing of a Textile Test Method That Produces Non-Normally Distributed Data
Standard Practice for Interlaboratory Testing of a Textile Test Method That Produces Non-Normally Distributed Data
SCOPE
1.1 This practice covers design and analysis of interlaboratory testing of a test procedure in the case where the resulting test data are discrete variates or are continuous variates not normally distributed. This practice applies to all such interlaboratory tests used to validate a test procedure.
1.2 Analysis of interlaboratory test results permits validation that the process of using the test method is in statistical control and provides the information required to write statements on precision and bias as directed in Practice D2906. It also gives the information for determining the number of specimens per unit in the laboratory sample as required in Practice D2905.
1.3 Precision statements for non-normally distributed data can be written as a function of the level of the property of interest without an interlaboratory test if the underlying distribution is known and statistical control can be assumed.
1.4 If the underlying distribution is unknown, the precision of the test method can only be approximated. There are no generally accepted methods of making approximations of this sort.
1.5 If statistical control cannot be assumed, then a meaningful precision statement cannot be written and the test method should not be used.
1.6 This practice is intended for use with data from test methods that cannot be properly modeled by a normal distribution. See Practices D2904 and E691 for applications that can be modeled by a normal distribution.
1.7 This practice includes the following sections: SectionsScope1Referenced Documents 2Terminology 3Significance and Uses 4General Considerations 5Basic Statistical Design 6Pilot-Scale Interlaboratory Test 7Full-Scale Interlaboratory Test 8Missing Data 9Outlying Observations 10Interpretation of Data 11Plotting Results 12Keywords 13Pilot-Scale and Full-Scale Interlaboratory Tests Annex A1Calculation of Chi-Square Annex A2
1.8 This standard does not purport to address all of the safety concerns, if any, associated with its use. It is the responsibility of whoever uses this standard to consult and establish appropriate safety and health practices and determine the applicability of regulatory limitations prior to use.
General Information
Relations
Standards Content (Sample)
NOTICE: This standard has either been superseded and replaced by a new version or discontinued.
Contact ASTM International (www.astm.org) for the latest information.
Designation: D 4467 – 94
Standard Practice for
Interlaboratory Testing of a Textile Test Method That
Produces Non-Normally Distributed Data
This standard is issued under the fixed designation D 4467; 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
Plotting Results 12
Keywords 13
1.1 This practice covers design and analysis of interlabora-
Pilot-Scale and Full-Scale Interlaboratory Tests Annex A1
tory testing of a test procedure in the case where the resulting
Calculation of Chi-Square Annex A2
test data are discrete variates or are continuous variates not
1.8 This standard does not purport to address all of the
normally distributed. This practice applies to all such interlabo-
safety concerns, if any, associated with its use. It is the
ratory tests used to validate a test procedure.
responsibility of whoever uses this standard to consult and
1.2 Analysis of interlaboratory test results permits valida-
establish appropriate safety and health practices and deter-
tion that the process of using the test method is in statistical
mine the applicability of regulatory limitations prior to use.
control and provides the information required to write state-
ments on precision and bias as directed in Practice D 2906. It
2. Referenced Documents
also gives the information for determining the number of
2.1 ASTM Standards:
specimens per unit in the laboratory sample as required in
D 123 Terminology Relating to Textiles
Practice D 2905.
D 2904 Practice for Interlaboratory Testing of a Textile Test
1.3 Precision statements for non-normally distributed data
Method that Produces Normally Distributed Data
can be written as a function of the level of the property of
D 2905 Practice for Statements on Number of Specimens
interest without an interlaboratory test if the underlying distri-
for Textiles
bution is known and statistical control can be assumed.
D 2906 Practice for Statements on Precision and Bias for
1.4 If the underlying distribution is unknown, the precision
Textiles
of the test method can only be approximated. There are no
D 4646 Test Method for 24-h Batch-Type Measurement of
generally accepted methods of making approximations of this
Containment Sorption by Soils and Sediments
sort. 4
D 4853 Guide for Reducing Test Variability
1.5 If statistical control cannot be assumed, then a mean-
E 456 Terminology Relating to Quality and Statistics
ingful precision statement cannot be written and the test
E 691 Practice for Conducting an Interlaboratory Study to
method should not be used.
Determine the Precision of a Test Method
1.6 This practice is intended for use with data from test 5
E 1169 Guide for Conducting Ruggedness Tests
methods that cannot be properly modeled by a normal distri-
bution. See Practices D 2904 and E 691 for applications that
3. Terminology
can be modeled by a normal distribution.
3.1 Definitions:
1.7 This practice includes the following sections:
3.1.1 test method, n—a definitive procedure for the identi-
Sections
fication, measurement, and evaluation of one or more qualities,
Scope 1
characteristics, or properties of a material, product, system, or
Referenced Documents 2
Terminology 3 service that produces a test result.
Significance and Uses 4
3.1.2 For definitions of textile and statistical terms used in
General Considerations 5
this practice and discussions of their use, refer to Terminology
Basic Statistical Design 6
Pilot-Scale Interlaboratory Test 7 D 123, and Terminology E 456.
Full-Scale Interlaboratory Test 8
3.2 Definitions of Terms Specific to This Standard:
Missing Data 9
3.2.1 assignable cause—a factor which contributes to varia-
Outlying Observations 10
Interpretation of Data 11
tion and is feasible to detect and identify.
3.2.2 interlaboratory testing—the evaluating of a test
1 2
This practice is under the jurisdiction of ASTM Committee D-13 on Textiles Annual Book of ASTM Standards, Vol 07.01.
and is the direct responsibility of Subcommittee D13.93 on Statistics. Annual Book of ASTM Standards, Vol 11.04.
Current edition approved June 15, 1994. Published August 1994. Originally Annual Book of ASTM Standards, Vol 07.02.
published as D 4467 – 85. Last previous edition D 4467 – 85. Annual Book of ASTM Standards, Vol 14.02.
Copyright © ASTM, 100 Barr Harbor Drive, West Conshohocken, PA 19428-2959, United States.
D 4467
method in more than one laboratory by analyzing data obtained 4.8 Interlaboratory tests of the type discussed in this prac-
from one or more materials that are as homogeneous as tice are used to locate and measure the sources of variability
practical. associated with a test method when the test method is used to
3.2.3 random cause—one of many factors which contribute evaluate a property of one or more materials, each of which is
to variation but which are not feasible to detect and identify as homogeneous as practical with respect to that property. Such
since they are random in origin and usually small in effect. interlaboratory tests provide no information about the sources
3.2.4 state of statistical control—a condition in which a of variability associated with the sampling of the stream of
process, including a measurement process, is subject only to product from a manufacturing process, a shipment, or material
random variation. in inventory. Estimation of such sampling errors requires an
entirely different type of experiment which is not specified
4. Significance and Use
presently in an ASTM Committee D-13 standard.
4.1 The planning of interlaboratory tests requires a general
5. General Considerations
knowledge of statistical principles. Interlaboratory tests should
be planned, conducted, and analyzed after consultation with 5.1 Overview—This section covers various aspects of allo-
statisticians who are experienced in the design and analysis of cating specimens to the participating laboratories.
experiments and who have some knowledge of the nature of 5.2 Sampling of Materials—Select a source of samples of
the variability likely to be encountered in the test method. material in such a way that any one portion of the material,
4.2 The instructions of this practice are specifically appli- within which laboratories, operators, days, and other factors
cable to the design and analysis of the following tests: are to be compared, will be as homogeneous as possible with
4.2.1 Pilot-scale interlaboratory tests and respect to the property being measured. Otherwise, increased
4.2.2 Full-scale interlaboratory tests. replication will be required to reduce the size of the difference
4.3 Procedures given in this practice are applicable to which can be detected.
methods based on the measurement of the following types of 5.3 Randomization of Specimens:
variates: 5.3.1 Complete Randomization—Randomize the selection
4.3.1 Ratings (grades or scores), such as those resulting of specimens for each laboratory sample; divide all the
from comparisons with AATCC gray scales, randomized specimens of a specific material, after labeling,
4.3.2 Percent of observations with a specific attribute, into the required number of groups, each group corresponding
4.3.3 Counts of attributes, such as number of nonconformi- to a specific laboratory.
ties, 5.3.2 Stratification—In some cases it is advantageous to
4.3.4 Any data not normally distributed which the analyst follow a stratified pattern in the allocations of the specimens to
cannot or prefers not to transform, such as flammability data or laboratories. For example, if the specimens are bobbins of yarn
percent extractables. from different spinning frames, it is better to allocate to each
4.4 Interlaboratory testing is a means of determining the laboratory equal numbers of specimens from each spinning
consistency of results when the same material is tested under frame. In such cases, the specimens within each spinning frame
varying conditions such as: operators, laboratories, equipment, are randomized separately rather than all of the specimens from
or environment. An interlaboratory test should do the follow- all of the frames.
ing: 5.4 Order of Tests—In many situations, variability among
4.4.1 Show if the test method distinguishes between levels replicate tests is greater when measurements are made at
of the property being tested, different times than when they are made together as part of a
4.4.2 Show if the test method is in statistical control; group. Sometimes trends are apparent among results obtained
statistical control being the presence of only random variation, consecutively. Furthermore, some materials undergo measur-
4.4.3 Detect operators, laboratories, and equipment out of able changes within relatively short storage periods. For these
statistical control. reasons, treat the dates of testing, as well as the order of tests
4.5 An initial single-laboratory preliminary test of a test carried out in a group as controlled, systematic variables.
procedure is necessary, usually including ruggedness testing, to 5.5 Selecting the Measure of Average Performance—Data
determine the feasibility of the method and to determine the are summarized for presentation and analysis by use of some
method’s sensitivity to variables which must be controlled. See measure of typical performance. For textile testing, there are
Guides D 4853 or E 1169 for a discussion of ruggedness usually three choices:
testing. 5.5.1 Arithmetic Average—The arithmetic average is the
4.6 A pilot-scale interlaboratory test may be needed to measure of choice when the data are symmetrically distributed
identify sources of variation, to establish clarity of instructions or are from a Poisson distribution.
of the proposed operating procedures, and to obtain estimates 5.5.2 Median—The median (midpoint, fiftieth percentile) is
as to the number of test results per operator per material to be the preferred measure when the data are asymmetrically
used in the initial full-scale interlaboratory test. distributed. When the distribution is symmetrical, the arith-
4.7 A full-scale interlaboratory test is usually made after a metic average and the median are equal.
pilot-scale test. If the task group prefers, a full-scale test may 5.5.3 Proportion—A proportion, which may be expressed as
be run without a previous pilot-scale test but with the under- a fraction (decimal) or percent, is the measure to use when the
standing that unsatisfactory results would require another data are counts of items having a particular attribute out of a
full-scale test. specified number of items.
D 4467
5.6 Number of Replicate Specimens—The number of speci- of determinations distributed over fewer materials. In the same
mens tested by each operator in each laboratory for each way, a specific number of determinations per material will
material may be calculated from previous information or from yield more information if they are spread over the largest
a pilot run. This number of specimens or replications (mini- number of laboratories possible. For the recommended mini-
mum of two) depends on the relative size of the random error mum design, see 6.2. If experience with the pilot-scale inter-
and the smallest effect to be detectable. A replicate consists of laboratory test casts doubt on the adequacy of the starting
one specimen of each condition and material to be tested in the design, estimate the number of determinations needed to detect
statistical design. the smallest differences of practical importance.
5.6.1 Symmetrical Non-Normal Distributions—Calculate 5.8 Multiple Equipment (Instruments)—When multiple in-
the number of observations required in each mean using Eq 1 struments within a laboratory are used on an interlaboratory
(Note 1): test, tests should be made on all equipment to establish the
presence or absence of the equipment effects. All types of
2 2
n 5 ~ts/E! 5 16~s/E! (1)
equipment allowed by a test method should be tested to allow
greatest flexibility. If an equipment effect is present and cannot
where:
n 5 number of observations in each mean, be eliminated by use of pertinent scientific principles, known
t5 4 5 specified value in Tchebychev’s inequality (Note
standards should be run and appropriate within-laboratory
2), quality control procedure should be used.
s 5 standard deviation for individual observations ob-
6. Basic Statistical Design
tained from previously conducted studies, and
E 5 smallest difference it is of practical importance to
6.1 It is advisable to keep the design as simple as possible,
detect, expressed in the same units of measure as the
yet to obtain estimates of within- and between-laboratory
averages and standard deviation.
variation unconfounded with secondary effects. Provisions also
should be made for estimates of significance of variation due
NOTE 1—With a balanced design, half of the total observations in the
to: materials-by-laboratories interactions, and operators-by-
experiment will be in each of the two sample means used to determine the
materials interactions.
possible effect of each factor being evaluated at two levels; one third of the
total observations will be in each of the three sample means used to
6.2 Include in the basic statistical design the following:
determine the possible effect of each factor being evaluated at three levels;
6.2.1 A minimum of three materials spanning the range of
and so on. The required value of n refers to such means.
interest for the property being measured,
NOTE 2—Tchebychev’s inequality states that in all cases at least
6.2.2 At least ten laboratories unless the test method cannot
(1−1/t ) of the total observations, n, will lie within the closed range x¯ 6
be used in that many laboratories,
ts , when t is not less than 1. For t5 4, at least 93.75 % of all
6.2.3 A recommended minimum of two operators per labo-
observations will fall within x¯ 6 4s. For symmetrical distributions, the
ratory, and
observed percentage is usually well above the minimum percentage
specified by Tchebychev’s inequality.
6.2.4 At least two specimens of each material to be tested by
each operator in a designated random order.
5.6.2 Asymmetrical Distribution Except Poisson or
6.3 The laboratory report format is presented in Table 1.
Binomial—Calculate the number of observations required in
6.4 Select materials to produce a wide range of expected
each mean using Eq 2 (Note 2):
results. The materials should include the applicable physical
2 2
n 5 ~1.25ts/E! 5 25~s/E! (2)
forms. For example, if woven fabric, knit fabric, and non-
wher
...
Questions, Comments and Discussion
Ask us and Technical Secretary will try to provide an answer. You can facilitate discussion about the standard in here.