You are on page 1of 10

An American National Standard

Designation: E 456 02e1

Standard Terminology
Relating to Quality and Statistics1
This standard is issued under the fixed designation E 456; 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.

e1 NOTEEditorial changes were made throughout in January 2004.

1. Scope its usage. These references may be to other practices and


1.1 This terminology includes those quality and statistical guides or to more specific terminology standards, such as
terms in wide use in ASTM for which standard definitions Terminology E 1325.
appear desirable. 4. Terminology
2. Referenced Documents acceptance (control chart or acceptance control chart
2.1 ASTM Standards: 2 usage, n), na decision that the process is operating in a
E 29 Practice for Using Significant Digits in Test Data to satisfactory manner with respect to the statistical measures
Determine Conformance with Specifications being plotted: action limits: control limits.
E 177 Practice for the Use of the Terms Precision and Bias accepted reference value, na value that serves as an
in ASTM Test Methods agreed-upon reference for comparison, and which is derived
E 178 Practice for Dealing with Outlying Observations as: (1) a theoretical or established value, based on scientific
E 1169 Guide for Conducting Ruggedness Tests principles, (2) an assigned or certified value, based on
E 1325 Terminology Relating to Design of Experiments experimental work of some national or international organi-
E 1402 Terminology Relating to Sampling zation, or (3) a consensus or certified value, based on
E 1488 Guide for Statistical Procedures to Use in Develop- collaborative experimental work under the auspices of a
ing and Applying Test Methods scientific or engineering group.
E 1994 Practice for Use of Process Oriented AOQL and accuracy, nthe closeness of agreement between a test result
LTPD Sampling Plans and an accepted reference value.
E 2281 Practice for Process and Measurement Capability NOTE 1The term accuracy, when applied to a set of test results,
Indices involves a combination of a random component and of a common
E 2282 Guide for Defining the Test Result of a Test Method systematic error or bias component.
E 2334 Practice for Setting an Upper Confidence Bound for
a Fraction or Number of Non-Conforming Items, or a Rate aliases, nin a fractional factorial design, two or more effects
of Occurrence for Non-Conformities, Using Attribute Data, which are estimated by the same contrast and which,
When there is a Zero Response in the Sample therefore, cannot be estimated separately. E 1325
assignable cause, na factor that contributes to variation, and
3. Significance and Use which is feasible to detect and identify.
3.1 This terminology is the general terminology standard for NOTE 2Many factors will contribute to variation but it may not be
terms defined by Committee E-11. feasible (economically or otherwise) to identify some of them.
3.2 Citation is made to other E-11 standards which contain
attribute data, nobserved values or determinations which
more extensive information regarding the particular term and
indicate the presence or absence of specific characteristics.
DISCUSSIONItems or units of material may be evaluated by counting
1
This terminology is under the jurisdiction of ASTM Committee E 11 on Quality or measurement. Attributes are counted whereas variables are mea-
and Statistics and is the direct responsibility of Subcommittee E11.60 on Terminol- sured. Attribute distributions are discrete. See variables data.
ogy.
Current edition approved Oct. 10, 2002. Published November 2002. Originally attributes, method of, nmeasurement of quality by the
published as E 456 72. Last previous edition E 456 96. method of attributes consists of noting the presence (or
2
For referenced ASTM standards, visit the ASTM website, www.astm.org, or
contact ASTM Customer Service at service@astm.org. For Annual Book of ASTM absence) of some characteristic or attribute in each of the
--`,,,````,,````,``,,```,,`,-`-`,,`,,`,`,,`---

Standardsvolume information, refer to the standards Document Summary page on units in the group under consideration, and counting how
the ASTM website.

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

Copyright ASTM International 1Document provided by IHS Licensee=Instituto Mexicano Del Petroleo/3139900100,
Reproduced by IHS under license with ASTM 10/27/2004 13:11:48 MDT Questions or comments about this message: please call
the Document Policy Group at 303-397-2295.
E 456 02e1
many units do (or do not) possess the quality attribute, or cartons that contain boxes or packages within them.
how many such events occur in the unit, group, or area.
completely randomized design, na design in which the
E 2334
treatments are assigned at random to the full set of experi-
average outgoing quality (AOQ)the average percent defec-
mental units. E 1325
tive of outgoing product including all accepted lots or
batches, after any defective units found in them are replaced completely randomized factorial design, na factorial ex-
by acceptable units, plus all lots or batches which are not periment (including all replications) run in a completely
accepted after such lots or batches have been effectively randomized design. E 1325
100 % inspected and all defective units replaced by accept- component of variance, na part of a total variance identified
able units. E 1994 with a specified source of variability.
average outgoing quality limit (AOQL)the maximum of composite design, na design developed specifically for
the AOQs for all possible incoming percentages defective for fitting second order response surfaces to study curvature,
the process, for a given acceptance sampling plan. E 1994 constructed by adding further selected treatments to those
average quality protectiona type of protection in which obtained from a 2n factorial (or its fraction). E 1325
there is prescribed some chosen value of average percent confidence bound, nsee confidence limit. E 2334
defective in the product after inspection (average outgoing confidence coefficient, nthe value, C, of the probability
quality limit (AOQL), that shall not be exceeded in the long associated with a confidence interval or statistical coverage
run no matter what may be the level of percent defective in interval. It is often expressed as a percentage. ISO 3534-1
the product submitted to the inspector. E 1994 E 2334
average run length (ARL)(1) sample sense, nthe aver- confidence level, nsee confidence coeffcient. E 2334
age number of times that a process will have been sampled confidence limit, neach of the limits, T1 and T2, of the two
and evaluated before a shift in process level is signaled, and sided confidence interval, or the limit T of the one sided
(2) unit sense, nthe average number of units that will have confidence interval. E 2334
been produced before a shift in level is signaled. confounded factorial design, na factorial experiment in
which only a fraction of the treatment combinations are run
DISCUSSIONA long ARL is desirable for a process located at its in each block and where the selection of the treatment
specified level (so as to minimize calling for unneeded investigation or combinations assigned to each block is arranged so that one
corrective action) and a short ARL is desirable for a process shifted to
some undesirable level (so that corrective action will be called for
or more prescribed effects is(are) confounded with the block
promptly). ARL curves are used to describe the relative quickness in effect(s), while the other effects remain free from confound-
detecting level shifts of various control chart systems. ing.
average standard deviation, s, narithmetic average of NOTE 5All factor level combinations are included in the experiment.
sample standard deviations. E 2281 E 1325
balanced incomplete block design (BIB), nan incomplete
confounding, ncombining indistinguishably the main effect
block design in which each block contains the same number
of a factor or a differential effect between factors (interac-
k of different versions from the t versions of a single
tions) with the effect of other factor(s), block factor(s) or
principal factor arranged so that every pair of versions
interactions(s).
occurs together in the same number, l, of blocks from the b
blocks. E 1325 NOTE 6Confounding is a useful technique that permits the effective
batch, na definite quantity of some product or material use of specified blocks in some experiment designs. This is accomplished
produced under conditions that are considered uniform. by deliberately preselecting certain effects or differential effects as being
of little interest, and arranging the design so that they are confounded with
NOTE 3A batch is usually smaller than a lot. block effects or other preselected principal factor or differential effects,
while keeping the other more important effects free from such complica-
bias, nthe difference between the expectation of the test tions. Sometimes, however, confounding results from inadvertent changes
results and an accepted reference value. to a design during the running of an experiment or from incomplete
planning of the design, and it serves to diminish, or even to invalidate, the
NOTE 4Bias is the total systematic error as contrasted to random
error. There may be one or more systematic error components contributing effectiveness of an experiment. E 1325
to the bias. A larger systematic difference from the accepted reference
value is reflected by a larger bias value.
consumers riskthe probability that a lot whose percentage
defective is equal to the LTPD will be accepted by the plan.
characteristic, na property of items in a sample or popula- E 1994
tion which, when measured, counted or otherwise observed, contrast, na linear function of the observations for which
helps to distinguish between the items. E 2282 the sum of the coefficients is zero.
cluster sampling, nwhen the primary sampling unit com-
NOTE 7With observations Y1, Y2,..., Yn, the linear function
prises a bundle of elementary units or a group of subunits, a1Y1 + a2Y2 + ... + an Yn is a contrast if, and only if (ai = 0, where the ai
the term cluster sampling may be applied.
values are called the contrast coefficients. E 1325
DISCUSSIONExamples of cluster sampling are: selection of city
blocks as primary sampling units; selection of a household as a cluster contrast analysis, na technique for estimating the param-
of people (of which only one may be interviewed); selection of bundles eters of a model and making hypothesis tests on preselected
of rods or pipe from a shipment; and selection, from a shipment, of linear combinations of the treatments (contrasts).

Copyright ASTM International


--`,,,````,,````,``,,```,,`,-`-`,,`,,`,`,,`---
2Document provided by IHS Licensee=Instituto Mexicano Del Petroleo/3139900100,
Reproduced by IHS under license with ASTM 10/27/2004 13:11:48 MDT Questions or comments about this message: please call
the Document Policy Group at 303-397-2295.
--`,,,````,,````,``,,```,,`,-`-`,,`,,`,`,,`---
E 456 02e1
NOTE 8Contrast analysis involves a systematic tabulation and analy- the process should be obtained along with a product, and that designed
sis format usable for both simple and complex designs. When any set of experiments using relatively small shifts in factor levels (within produc-
orthogonal contrasts is used, the procedure, as in the example, is tion tolerances) can yield this knowledge at minimum cost. The range of
straightforward. When terms are not orthogonal, the orthogonalization variation of the factors for any one EVOP experiment is usually quite
process to adjust for the common element in nonorthogonal contrast is small in order to avoid making out of tolerance products, which may
also systematic and can be programmed. E 1325 require considerable replication, in order to be able to clearly detect the
effect of small changes. E 1325
control(evaluation), nan evaluation to check, test, or
verify; (authority): the act of guiding, directing, or manag- experimental design, nsee design of experiments. E 1325
ing; (stability): a state of process in which the variability is experiment space, nthe materials, equipment, environmen-
attributable to a constant system of chance causes. tal conditions and so forth that are available for conducting
control chart factor, na factor, usually varying with sample an experiment. E 1325
size, to convert specified statistics or parameters into a experimental unit, na portion of the experiment space to
central line value or control limit appropriate to the control which a treatment is applied or assigned in the experiment.
chart.
control chart method, nthe method of using control charts NOTE 13The unit may be a patient in a hospital, a group of animals,
to determine whether or not processes are in a stable state. a production batch, a section of a compartmented tray, etc. E 1325
control limits, nlimits on a control chart which are used as
factorial experiment (general), nin general, an experiment
criteria for signaling the need for action, or for judging
in which all possible treatments formed from two or more
whether a set of data does or does not indicate a state of
factors, each being studied at two or more levels (versions)
statistical control.
are examined so that interactions (differential effects) as well
conventional true value of a quantity, nvalue attributed to
as main effects can be estimated. E 1325
a particular quantity and accepted, sometimes by conven-
tion, as having an uncertainty appropriate for a given 2n factorial experiment, na factorial experiment in which n
purpose. factors are studied, each of them in two levels (versions).
E 1325
NOTE 988Conventional true value is sometimes called 88assigned fractional factorial design, na factorial experiment in
value, 88best value, 88conventional value, or 88reference value. 88Ref-
which only an adequately chosen fraction of the treatments
erence value, in this sense, should not be confused with 88reference
value in the sense of an influence quantity affecting a measuring required for the complete factorial experiment is selected to
instrument. be run.
NOTE 10Frequently, a number of results of measurements of a
NOTE 14This procedure is sometimes called fractional replication.
quantity is used to establish a conventional true value.
DISCUSSIONWhen warning limits are used, the control limits are E 1325
often called action limits. Action may be in the form of investigation
of the source(s) of an assignable cause, making a process adjustment, frame, na list, compiled for sampling purposes, which
or terminating a process. Criteria other than control limits are also used designates the items (units) of a population or universe to be
frequently. considered in a study.
dependent variable, nSee response variable. E 1325 DISCUSSIONWhen a frame is available, sampling schemes can be
design of experiments, nthe arrangement in which an devised for selection of the units directly (one-stage), or in two or more
experimental program is to be conducted, and the selection stages. In multi-stage sampling, a frame is needed for each stage. As an
of the levels (versions) of one or more factors or factor example, the cartons of a lot could be the first-stage units, packages
within the carton could be second-stage units, and items within the
combinations to be included in the experiment. Synonyms packages could be the third-stage units.
include experiment design and experimental design.
E 1325 fully nested experiment, na nested experiment in which the
deviation, nthe difference between a measurement or quasi- second factor is nested within levels (versions) of the first
measurement and its stated value or intended level. factor and each succeeding factor is nested within versions
DISCUSSIONDeviation should be stated as a difference in terms of
of the previous factor. E 1325
the appropriate data units. Sometimes these units will be original hierarchical experiment, nsee nested experiment.
measurement units; sometimes they will be quasi-measurements; that E 1325
is, a scaled rating of subjective judgments; sometimes they will be incomplete block design, na design in which the experi-
designated values representing all continuous or discrete measurements ment space is subdivided into blocks in which there are
falling in defined cells or classes.
insufficient experimental units available to run a complete
error of result, nthe test result minus the accepted reference set of treatments or replicate of the experiment. E 1325
value (of the characteristic). intermediate precisions, nthe closeness of agreement be-
NOTE 11It is not possible to correct for random error. tween test results obtained under specified intermediate
precision conditions.
evolutionary operation (EVOP), na sequential form of
experimentation conducted in production facilities during NOTE 15The specific measure and the specific conditions must be
regular production. specified for each intermediate measure of precision; thus, 88standard
deviation of test results among operators in a laboratory, or 88day-to-day
NOTE 12The principal theses of EVOP are that knowledge to improve standard deviation within a laboratory for the same operator.

Copyright ASTM International 3Document provided by IHS Licensee=Instituto Mexicano Del Petroleo/3139900100,
Reproduced by IHS under license with ASTM 10/27/2004 13:11:48 MDT Questions or comments about this message: please call
the Document Policy Group at 303-397-2295.
E 456 02e1
NOTE 16Because the training of operators, the agreement of different limit. E 2281
pieces of equipment in the same laboratory and the variation of environ- lower tolerance limit (LTL) (lower specification limit), na
mental conditions with longer time intervals all depend on the degree of tolerance limit that defines the lower conformance boundary
within-laboratory control, the intermediate measures of precision are
likely to vary appreciably from laboratory to laboratory. Thus, intermedi-
for an individual unit of a manufacturing or service opera-
ate precisions may be more characteristic of individual laboratories than of tion.
the test method. main effect, average effect, na term describing a measure
for the comparison of the responses at each level (version) of
intermediate precision conditions, nconditions under a factor averaged over all levels (versions) of other factors in
which test results are obtained with the same test method the experiment.
using test units or test specimens (see Practice E 691,2 10.3)
taken at random from a single quantity of material that is as NOTE 18The term 88main effect may describe the parameter in an
nearly homogeneous as possible, and with changing condi- assumed model or the estimate of this parameter. E 1325
tions such as operator, measuring equipment, location within method of least squares, na technique of estimation of a
the laboratory, and time. parameter which minimizes (e2, where e is the difference
item, n(1) an object or quantity of material on which a set of between the observed value and the predicted value derived
observations can be made: (2) an observed value or test from the assumed model. E 1325
result obtained from an object or quantity of material. minimum process capability index, Cpk, nsmaller of the
DISCUSSIONThe second usage in the definition is generally limited upper process capability index and the lower process capa-
to generic descriptions such as in the definition of population. Terms bility index. E 2281
such as observation, measurement, test result, unit, value or minimum process performance index, Ppk, nsmaller of the
yield are more common in specific applications. A set as used here upper process performance index and the lower process
may be one or more variables. performance index. E 2281
level (of a factor), na given value, a specification of mixture design, na design in which two or more ingredients
procedure or a specific setting of a factor. or components shall be mixed and the response is a property
of the resulting mixture that does not depend upon the
NOTE 1788Version is a general term applied both to quantitative and amount of the mixture.
qualitative factors. The more restrictive term 88level is frequently used to
express more precisely the quantitative characteristic. For example, two NOTE 19The proportions of each of the q components (Xi) in the
versions of a catalyst may be presence and absence. Four levels of a heat c
treatment may be 100C, 120C, 140C, and 160C. E 1325 mixture shall satisfy the conditions O # Xi # 1 and ( Xi = 1; and each
i51
experimental point is defined in terms of these proportions.
long term standard deviation, sLT, nsample standard NOTE 20In some fields of application the experimental mixtures are
deviation of all individual (observed) values taken over a described by the terms 88formulation or 88blend. The use of mixture
long period of time. designs is appropriate for experimenting with the formulations of manu-
factured products, such as paints, gasoline, foods, rubber, and textiles.
DISCUSSIONA long period of time may be defined as shifts, weeks,
NOTE 21In some applications, the proportions of the components of
or months, etc. E 2281 the mixture may vary between 0 and 100 % of the mixture (88complete
domain). In others, there may be operative restraints, so that at least one
lota definite quantity of a product or material accumulated
under conditions that are considered uniform for sampling component cannot attain 0 or 100 % (88reduced domain). E 1325
purposes. natural process limits (NPL), nlimits which include a
lot quality protectiona type of protection in which there is stated fraction of the individuals in a population.
prescribed some chosen value of limiting percent defective
NOTE 22Natural process limits will not ordinarily be the dimensional
in a lot (lot tolerance percent defective, (LTPD)) and also limits shown on an engineering drawing. They are mostly used to compare
some chosen value for the probability (called the consumers the natural capability of the process to tolerance limits.
risk) of accepting a submitted lot that has a percent defective DISCUSSIONFor populations with a normal (Gaussian) distribution,
equal to the lot tolerance percent defective. E 1994 the natural process limits ordinarily will be at 63 s. If placed around
lot tolerance percent defective (LTPD)the percentage of the standard level, these limits identify the boundaries which will
defective units in a batch or lot for which, for purposes of include approximately 99.7 % of the individuals in a process that is
acceptance sampling, the consumer wishes the probability of properly centered and in a state of statistical control. In many
circumstances (several machines making the same product that serially
acceptance to be restricted to a specified low value, specifi- feed into the process) it is recognized that in addition to the variability
cally 10 % for this practice. This is also referred to by the around a single level, an acceptable zone of standard levels (for the
more general term limiting quality taken at 10 % consumer different machines) is required. Then the NPL may be placed around
risk. E 1994 the Acceptable Process Levels (APL) that define this zone so that the
lower control limit (LCL), ncontrol limit for points below NPL identify the boundaries within which at least 99.7 % of the
the central line. individuals will be included in a process located at the APL, or inside
the zone. It should be noted that there is no assumption made that the
lower process capability index, Cpkl, nindex describing
process levels within the zone are random variables.
process capability in relation to the lower specification limit.
E 2281 nested experiment, nan experiment to examine the effect of
lower process performance index, Ppkl, nindex describing two or more factors in which the same level (version) of a
process performance in relation to the lower specification factor cannot be used with all levels (versions) of other
--`,,,````,,````,``,,```,,`,-`-`,,`,,`,`,,`---

Copyright ASTM International 4Document provided by IHS Licensee=Instituto Mexicano Del Petroleo/3139900100,
Reproduced by IHS under license with ASTM 10/27/2004 13:11:48 MDT Questions or comments about this message: please call
the Document Policy Group at 303-397-2295.
E 456 02e1
factors. Synonym: hierarchical experiment. E 1325 Plackett-Burman designs, na set of screening designs using
non-conforming item, nan item containing at least one orthogonal arrays that permit evaluation of the linear effects
non-conformity. of up to n = t 1 factors in a study of t, treatment
combinations. E 1325
DISCUSSIONThe term defective item is also used in this context.
population, nthe totality of items or units of material under
E 2334
consideration.
observation, n(1) the process of obtaining information DISCUSSIONThe word items may be interpreted in the sense of
regarding the presence or absence of an attribute of a test measurements, or possible measurements, for a single characteristic, or
specimen, or of making a reading on a characteristic or occasionally for multiple characteristics, on all items or units of
dimension of a test specimen, or (2) the attribute or mea- material being considered. The word totality may refer to items not
surement information obtained from the process. (The term available for inclusion in samples as well as those which are available.
88observed value is preferred for this second usage.) E 2334

NOTE 23See Annex A1. E 2282 precision, nthe closeness of agreement between independent
test results obtained under stipulated conditions.
observation, nthe process of obtaining information regard-
NOTE 28Precision depends on random errors and does not relate to
ing the presence or absence of an attribute of a test specimen,
the true value or the specified value.
or of making a reading on a characteristic or dimension of a NOTE 29The measure of precision usually is expressed in terms of
test specimen. imprecision and computed as a standard deviation of the test results. Less
NOTE 24Observation is also associated with the attribute or measure- precision is reflected by a larger standard deviation.
--`,,,````,,````,``,,```,,`,-`-`,,`,,`,`,,`---

ment information obtained from the process. The term observed value is NOTE 3088Independent test results means results obtained in a
manner not influenced by any previous result on the same or similar test
preferred for this second usage. E 2281 object. Quantitative measures of precision depend critically on the
observed value, nthe value obtained by carrying out the stipulated conditions. Repeatability and reproducibility conditions are
particular sets of extreme stipulated conditions.
complete protocol of the test method once, being either a
single test determination or an average or other specified probability sample, na sample of which the sampling units
combination of a specified number of test determinations. have been selected by a chance process. At each step of
selection, a specified probability of selection can be attached
NOTE 25See Annex A1. E 2282 to each sampling unit available for selection. E 1402
observed value, nthe value obtained by making an obser- probability sample, na sample of which the sampling units
vation. E 2281 have been selected by a chance process such that, at each
orthogonal array, na table of coefficients identifying the step of selection, a specified probability of selection can be
levels, or some weight associated with the levels, for each attached to each sampling unit available for selection.
factor to be used in the analysis of specified effects, which NOTE 31These probabilities of selection need not be equal. If equal,
are arranged in such a manner that each effect will be see simple random sample. See the general termsample. Also, see
independent of the other effects. E 1325 Practice E 1052 in this volume.
orthogonal contrasts, ntwo contrasts are orthogonal if the
process capability, PC, nstatistical estimate of the outcome
contrast coefficients of the two sets satisfy the condition that,
of a characteristic from a process that has been demonstrated
when multiplied in corresponding pairs, the sum of the
to be in a state of statistical control. E 2281
products is equal to zero. See contrast and contrast analy-
process capability index, Cp, nan index describing process
sis. E 1325
capability in relation to specified tolerance. E 2281
outliersee outlying observation. E 178
process performance, PP, nstatistical measure of the out-
outlying observation, nan observation that appears to devi-
come of a characteristic from a process that may not have
ate markedly in value from other members of the sample in
been demonstrated to be in a state of statistical control.
which it appears. E 178
E 2281
partially balanced incomplete block design (PBIB), nan
process performance index, Pp, nindex describing process
incomplete block design in which each block contains the
performance in relation to specified tolerance. E 2281
same number k, of different versions from the t versions of
random error of result, na component of the error which,
the principal factor.
in the course of a number of test results for the same
NOTE 26The arrangement is such that not all pairs of versions occur characteristic, varies in an unpredictable way.
together in the same number of the blocks; some versions can therefore be randomization, nthe procedure used to allot treatments at
compared with greater precision than others. E 1325 random to the experimental units so as to provide a high
partially nested experiment, na nested experiment in degree of independence in the contributions of experimental
which several factors may be crossed as in factorial experi- error to estimates of treatment effects.
ments and other factors nested within the crossed combina- NOTE 32An essential element in the design of experiments is to
tions. provide estimates of effects free from biases due to undetected assignable
causes within the experimental space. Randomization is a process to
NOTE 27It is not unusual to find that experiments consist of both minimize this risk. The operational procedure for assignment 88at random
factorial and nested segments. See nested experiment. E 1325 involves the use of random numbers or some similar method for assuring

Copyright ASTM International 5Document provided by IHS Licensee=Instituto Mexicano Del Petroleo/3139900100,
Reproduced by IHS under license with ASTM 10/27/2004 13:11:48 MDT Questions or comments about this message: please call
the Document Policy Group at 303-397-2295.
E 456 02e1
that each unit has an equal chance of being selected for each treatment. within-laboratory standard deviation. (See within-laboratory standard
E 1325 deviation.)

randomized block design, na design in which the experi- reproducibility, nprecision under reproducibility condi-
tions.
--`,,,````,,````,``,,```,,`,-`-`,,`,,`,`,,`---

ment space is subdivided into blocks of experimental units,


the units within each block being more homogeneous than reproducibility conditions, nconditions where test results
units in different blocks. are obtained with the same method on identical test items in
different laboratories with different operators using different
NOTE 33In each block the treatments are allocated randomly to the
experimental units within each block. Replication is obtained by the use of
equipment.
two or more blocks, depending on the precision desired, and a separate DISCUSSIONIdentical material means either the same test units or
randomization is made in each block. E 1325 test specimens are tested by all the laboratories as for a nondestructive
test or test units or test specimens are taken at random from a single
randomized block factorial design, na factorial experiment quantity of material that is as nearly homogeneous as possible. (See
run in a randomized block design in which each block Practice E 691.)
includes a complete set of factorial combinations. E 1325 DISCUSSIONA different laboratory of necessity means a different
range, R, nthe largest observation minus the smallest operator, different equipment, and different location and under different
observation in a set of values or observations. E 2281 supervisory control.
repeatability, nprecision under repeatability conditions.
reproducibility limit, n(R) the value below which the
NOTE 34Repeatability is one of the concepts or categories of the absolute difference between two test results obtained under
precision of a test method. reproducibility conditions may be expected to occur with a
NOTE 35Measures of repeatability defined in this compilation are probability of approximately 0.95 (95 %).
repeatability standard deviation and repeatability limit.
NOTE 41The reproducibility limit is 2.8 ('1.96 =2 ) times the
repeatability conditions, nconditions where independent reproducibility standard deviation. The multiplier is independent of the
test results are obtained with the same method on identical size of the interlaboratory study (that is, of the number of laboratories
test items in the same laboratory by the same operator using participating), as explained in Practice E 177.2
the same equipment within short intervals of time. NOTE 42The approximation to 0.95 is reasonably good (say 0.90 to
0.98) when many laboratories (30 or more) are involved but is likely to be
NOTE 36See precision Note 3. poor when fewer than eight laboratories are studied.
DISCUSSIONThe same operator, same equipment requirement
means that for a particular step in the measurement process, the same reproducibility standard deviation (SR), nthe standard
combination of operator and equipment is used for every test result. deviation of test results obtained under reproducibility con-
Thus, one operator may prepare the test specimens, a second measure ditions.
the dimensions and a third measure the mass in a test method for
determining density. NOTE 43Other measures of the dispersion of test results obtained
DISCUSSIONBy in the shortest practical period of time is meant under reproducibility conditions are the 88reproducibility variance and
that the test results, at least for one material, are obtained in a time the 88reproducibility coefficient of variation.
period not less than in normal testing and not so long as to permit NOTE 44The reproducibility standard deviation includes, in addition
significant change in test material, equipment or environment. to between-laboratory variability, the repeatability standard deviation and
a contribution from the interaction of laboratory factors (that is, differ-
repeatability limit (r), nthe value below which the absolute ences between operators, equipment and environments) with material
difference between two individual test results obtained under factors (that is, the differences between properties of the materials other
repeatability conditions may be expected to occur with a than that property of interest).
probability of approximately 0.95 (95 %).
residual error, nthe difference between the observed result
NOTE 37The repeatability limit is 2.8 ('1.96 =2 ) times the and the predicted value (estimated treatment response);
repeatability standard deviation. This multiplier is independent of the size Observed Result minus Predicted Value. E 1325
of the interlaboratory study, as explained in Practice E 177.2 response surface, nthe pattern of predicted responses based
NOTE 38The approximation to 0.95 is reasonably good (say 0.90 to on the empirical model derived from the experiment obser-
0.98) when many laboratories (30 or more) are involved, but is likely to
be poor when fewer than eight laboratories are studied.
vations. E 1325
response variable, nthe variable that shows the observed
repeatability standard deviation, nthe standard deviation results of an experimental treatment. Synonym dependent
of test results obtained under repeatability conditions. variable. E 1325
NOTE 39It is a measure of the dispersion of the distribution of test
robustness, ninsensitivity of a statistical test to departures
results under repeatability conditions. from underlying assumptions.
NOTE 40Similarly, 88repeatability variance and 88repeatability coef- DISCUSSIONMany statistical test procedures depend on the form of
ficient of variation could be defined and used as measures of the the assumed distribution of the population sampled to obtain exact
dispersion of test results under repeatability conditions. values for the probability statements. If departures from the assumed
DISCUSSIONIn an interlaboratory study, this is the pooled standard distribution do not materially affect the decisions which would be based
deviation of test results obtained under repeatability conditions. See on the statistical tests involved, the test is considered robust. For
Practice E 691. example, tests based on an assumption of normality that compare
DISCUSSIONThe repeatability standard deviation, usually consid- averages generally are robust even though the underlying distribution
ered a property of the test method, will generally be smaller than the of individual items in the population is not normal. On the other hand,

Copyright ASTM International 6Document provided by IHS Licensee=Instituto Mexicano Del Petroleo/3139900100,
Reproduced by IHS under license with ASTM 10/27/2004 13:11:48 MDT Questions or comments about this message: please call
the Document Policy Group at 303-397-2295.
E 456 02e1
the F-statistic for comparing variances may be an indicator of lack of 0.10 estimation of the precision of an estimator should take account of this
normality rather than a simple variance comparison. magnitude of f. See finite population correction.
ruggedness, ninsensitivity of a test method to departures sampling with replacement, na procedure used with some
from specified test or environmental conditions. probability sampling plans in which a selected unit is
DISCUSSIONAn evaluation of the ruggedness of a test method or
replaced after any step in selection so that this sampling unit
an empirical model derived from an experiment is useful in determining is available for selection again at the next step of selection,
whether the results or decisions will be relatively invariant over some or at any other succeeding step of the sample selection
range of environmental variability under which the test method or the procedure.
model is likely to be applied. E 1169 screening design, na balanced design, requiring relatively
minimal amount of experimentation, to evaluate the lower
ruggedness test, na planned experiment in which environ-
order effects of a relatively large number of factors in terms
mental factors or test conditions are deliberately varied in
of contributions to variability or in terms of estimates of
order to evaluate the effects of such variation.
parameters for a model.
DISCUSSIONSince there usually are many environmental factors that
NOTE 46In screening designs, the term lower order effects is some-
might be considered in a ruggedness test, it is customary to use a
times limited to first order terms such as linear components of main
screening type of experiment design (see screening design) which
effects, but often includes both first order terms and second order terms
concentrates on examining many first order effects and generally
such as two factor interactions and quadratic curvature components of
assume that second order effects such as interactions and curvature are
relatively negligible. Often in evaluating the ruggedness of a test main effects. E 1325
method, if there is an indication that the results of a test method are
short term standard deviation, sST, nthe inherent variation
highly dependent on the levels of the environmental factors, there is a
sufficient indication that certain levels of environmental factors must be present when a process is operating in a state of statistical
included in the specifications for the test method, or even that the test control, expressed in terms of standard deviation.
method itself will need further revision. E 1169 DISCUSSIONThis may also be stated as the inherent process varia-
run, n(1) an uninterrupted sequence of occurrences of the tion. E 2281
same attribute or event in a series of observations, and (2) a significant digit, nany of the figures 0 through 9, except
consecutive set of successively increasing run-up or succes- leading zeros and some trailing zeros, which is used with its
sively decreasing run-down values in a series of variable place value to denote a numerical quantity to some desired
measurements. approximation
DISCUSSIONIn control chart applications, some variable measure- NOTE 47The digit zero may either indicate a specific value or indicate
ments are treated as attributes in determining runs. For example, a run place only. Zeros leading the first nonzero digit of a number indicate order
might be considered a series of a specified number of consecutive of magnitude only and are not significant digits. For example, the number
points above or below the central line.3 0.0034 has two significant digits. Zeros trailing the last nonzero digit for
numbers represented with a decimal point are significant digits. For
sample, na group of items, observations, test results, or
example, the numbers 1270. and 32.00 each have four significant digits.
portions of material, taken from a large collection of items, The significance of trailing zeros for numbers represented without use of
observations, test results, or quantities of material, which a decimal point can only be identified from knowledge of the source of the
serves to provide information that may be used as a basis for value. For example, a modulus strength, stated as 140 000 Pa, may have
making a decision concerning the larger collection. as few as two or as many as six significant digits.
To eliminate ambiguity, the exponential notation may be used. Thus,
DISCUSSIONThe sample may be the units of material themselves or 1.40 3 105 indicates that the modulus is reported to the nearest 0.01 3 10
the set of the observations collected from them. The decision may or 5 or 1000 Pa.
may not involve taking action on the units of material, or on the Use of appropriate SI prefixes is recommended for metric units to
process. It is necessary to describe whether the sample is to be selected reduce the need for trailing zeros of uncertain significance. Thus, 140 kPa
on a simple random, a stratified random, or other specified basis. and 0.140 MPa each indicate that the modulus is reported to the nearest 1
Probability samples, that is, samples selected by chance using appro- kPa or 1000 Pa, while 140 kPa may again have two or three significant
priate randomization, are required to make confidence interval state-
ments and similar statistical inferences about the parameters of the
digits. E 29
sampled population. E 2334 special cause, nsource of intermittent variation in a process.
sample size, nthe number of units in a sample or the number ISO 3534-2
of observations in a sample. E 2334 DISCUSSIONSometimes special cause is taken to be synonymous
sampling fraction, f, nthe ratio f of the number of sampling with assignable cause. However a distinction should be recognized.
units selected for the sample to the number of sampling units A special cause is assignable only when it is specifically identified. Also
available. a common cause may be assignable.
DISCUSSIONA special cause arises because of specific circumstances
NOTE 45For the simple random sample case, f = n/N where n is the which are not always present. As such, in a process subject to special
sample size and N is the number of sampling units available. When f > causes, the magnitude of the variation from time to time is unpredict-
able. E 2281
3
Other examples may be found in references such as Nelson, L. S., Interpreting
specification limits, nsee tolerance limits.
Shewhart X Control Charts, Journal of Quality Technology, Vol 17, No. 2, April stable process, nprocess in a state of statistical control;
1985. process condition when all special causes of variation have
--`,,,````,,````,``,,```,,`,-`-`,,`,,`,`,,`---

Copyright ASTM International 7Document provided by IHS Licensee=Instituto Mexicano Del Petroleo/3139900100,
Reproduced by IHS under license with ASTM 10/27/2004 13:11:48 MDT Questions or comments about this message: please call
the Document Policy Group at 303-397-2295.
E 456 02e1
been removed. ISO 3534-2 developed for non-integer N/k to overcome sampling bias. See Jessen.4
Estimation of the precision of an average computed from a
DISCUSSIONObserved variation can then be attributed to random systematic sample is a difficult problem that has no generally satisfac-
(common) causes. Such a process will generally behave as though the tory solution. Independent replicate systematic samples provide an
results are simple random samples from the same population. approach to variance estimation, but have been rejected by some
DISCUSSIONThis state does not imply that the random variation is writers. In some ASTM situations where replicate samples may be
large or small, within or outside of specification, but rather that the obtained on a routine basis, the technique may be useful. See Cochran5
variation is predictable using statistical techniques. for an extended discussion of variance estimation for systematic
DISCUSSIONThe process capability of a stable process is usually sampling.
improved by fundamental changes that reduce or remove some of the
--`,,,````,,````,``,,```,,`,-`-`,,`,,`,`,,`---

test determination, n(1) the process of deriving from one or


random causes present and/or adjusting the mean towards the preferred
value.
more test observations (observed values) the presence or
DISCUSSIONContinual adjustment of a stable process will increase
absence of an attribute or the value of a characteristic or
dimension of a single test specimen, or (2) the attribute
variation. E 2281
(presence or absence) or value derived from the process (see
staggered nested experiment, na nested experiment in test specimen).
which the nested factors are run within only a subset of the NOTE 49See Annex A1.
versions of the first or succeeding factors. E 1325
standard deviation, nthe most usual measure of the disper- test determination, nthe value of a characteristic or dimen-
sion of observed values or results expressed as the positive sion of a single test specimen derived from one or more
square root of the variance. observed values. E 2282
test method, na definitive procedure that produces a test
statistic, na quantity calculated from a sample of observa-
result.
tions, most often to form an estimate of some population
parameter. DISCUSSIONExamples of test methods include, but are not limited
statistical measure, nstatistic or mathematical function of a to: identification, measurement, and evaluation of one or more quali-
ties, characteristics, or properties. [ASTM Regulations 2.2.6]
statistic.
E 2282
DISCUSSIONThe word statistical emphasizes that measures are
subject to inherent errors and that, in estimating a population parameter,
test observation, nsee observation.
they represent a sample, with inherent sampling variability. test result, nthe value of a characteristic obtained by
carrying out a specified test method.
statistical procedures, nthe organized techniques and meth-
NOTE 50The test method should specify that one or a number of
ods used to collect, analyze, and interpret data. individual observations be made and their average or another appropriate
DISCUSSIONStatistical procedures include the sampling consider- function, such as the medium or the standard deviation, be reported as the
ations or the experiment design for the collection of data, or both, and test result. It also may require standard corrections to be applied, such as
the numerical and graphical approaches to summarize and analyze the correction of gas volumes to standard temperature and pressure. A test
result, therefore, can be a result calculated from several observed values.
collected data. E 1488 In the simple case, the test result is the observed value itself. E 2282
subgroup, n(1) object sense, na set of units or quantity of test specimen, nthe portion of a test unit needed to obtain a
material obtained by subdividing a larger group of units or single test determination.
quantity of material, and (2) measurement sense, na set
of groups of observations obtained by subdividing a larger NOTE 51When used for a physical test, this is sometimes called 88test
piece. For a chemical test, it is sometimes called test portion or test
group of observations. See rational subgroup.
sample. For optical and other tests, it is also sometimes called test sample.
systematic error of result, na component of the error, In interlaboratory evaluation of test methods and other statistical proce-
which in the course of a number of test results for the same dures, it is best to reserve the word sample for the whole amount of
characteristic, remains constant or varies in a predictable material involved and not the individual test specimens, pieces or portions
way. being tested.
NOTE 52See Annex A1. E 2282
NOTE 48Systematic errors and their causes may be known or un-
known. test unit, nthe total quantity of material (containing one or
more test specimens) needed to obtain a test result as
systematic sampling, nsample selection procedure in which specified in the test method. See test result. E 2282
every kth element is selected from the universe or popula- tolerance limits (specification limits), nlimits that define
tion; for example, u, u + k, u + 2k, u + 3k, etc., where u is in the conformance boundaries for an individual unit of a
the interval 1 to k. manufacturing or service operation.
DISCUSSIONIf k = 20 and u = 7 is the initial unit selected, then DISCUSSIONLimits may be established either with or without the use
sampling units 7, 27, 47, 67, ..., would comprise the sample. When N/k
is not an integer, there is a small bias due to the end effect. When u is
selected by a chance process and N/k is an integer, the systematic 4
Jessen, R. J., Statistical Survey Techniques, John Wiley & Sons, Inc., New
sample will provide unbiased estimates of the population average or York, 1978, Sec. 12.2.
total. Situations for which N/k is not an integer usually ignore the small 5
Cochran, W. G., Sampling Techniques, John Wiley & Sons, Inc., New York,
or negligible bias in estimating the mean or total. Schemes have been 1977, Chapter 8.

Copyright ASTM International 8Document provided by IHS Licensee=Instituto Mexicano Del Petroleo/3139900100,
Reproduced by IHS under license with ASTM 10/27/2004 13:11:48 MDT Questions or comments about this message: please call
the Document Policy Group at 303-397-2295.
E 456 02e1
of probability considerations. Tolerance limits may be in the form of a upper control limit (UCL), ncontrol limit for points above
single (unilateral) limit (upper or lower) or double (bilateral) limits the central line.
(upper and lower). Double, or two-sided limits occur more frequently. upper process capability index, Cpku, nindex describing
Double limits are often stated as a symmetrical deviation from a stated
value, but they need not be symmetrical. Frequently the term specifi-
process capability in relation to the upper specification limit.
cation limits is used instead of tolerance limits. While tolerance limits E 2281
is generally preferred in terms of evaluating the manufacturing or upper process performance index (Ppku), nindex describ-
service requirements, specification limits may be more appropriate for ing process performance in relation to the upper specification
categorizing material, product, or service in terms of their stated limit. E 2281
requirements. upper tolerance limit (UTL) (upper specification limit),
tolerance specification, nthe total allowable variation na tolerance limit applicable to the upper conformance
around a level or state (upper limit minus lower limit), or the boundary for an individual unit of a manufacturing or service
maximum acceptable excursion of a characteristic. operation.
variables, method of, nmeasurement of quality by the
DISCUSSIONThe determination of the amount of variation to be method of variables consists of measuring and recording the
allowed involves the product or service requirements and consideration
numerical magnitude of a quality characteristic for each of
of process capability (see natural process limits), measurement
variability, and other appropriate elements or some compromise among the units in the group under consideration.
these. NOTE 57This involves reference to a continuous scale of some kind.
treatment, na combination of the levels (versions) of each variables data, nmeasurements which vary and may take
of the factors assigned to an experimental unit, synonym any of a specified set of numerical values.
treatment combination. E 1325
DISCUSSIONThe term random variable or variate is often used
treatment combination, nsee treatment. E 1325
to indicate that each of the specified set of values is associated with a
trueness, nthe closeness of agreement between the popula- specified relative frequency or probability, and that each is a random
tion mean of the measurements or test results and the sample from a continuous or a discrete, or discontinuous, population
accepted reference value. encompassing the specified values.
NOTE 53The measure of trueness usually is expressed in terms of variance, na measure of the squared dispersion of observed
bias. Greater bias means less favorable trueness. values or measurements expressed as a function of the sum
NOTE 5488Population mean is, conceptually, the average value of an of the squared deviations from the population mean or

--`,,,````,,````,``,,```,,`,-`-`,,`,,`,`,,`---
indefinitely large number of test results.
sample average.
NOTE 55Trueness is the systematic component of accuracy.
NOTE 58The sample variance, or variance of a sample of n observed
uncertainty, nan indication of the variability associated with values, is computed as s2 = [1/(n 1)][((yi y)2]. The sample standard
a measured value that takes into account two major compo- deviation s is the positive square root of the sample variance. The
nents of error: (1) bias, and (2) the random error attributed to population variance s2 = *R (y )2f (y)dy, where R is the region over
the imprecision of the measurement process. which the random variable y is defined, and where f (y) is the probability
density function and is the population mean of y. The population
DISCUSSIONQuantitative measures of uncertainty generally require standard deviation (s) is the positive square root of the population
descriptive statements of explanation because of differing traditions of variance.
usage and because of differing circumstances. For example: (1) the bias DISCUSSIONA listing of the sample variance s2 should always be
and imprecision may both be negligible; (2) the bias may not be accompanied by the degrees of freedom on which it is based. The
negligible while the imprecision is negligible; (3) neither the bias nor degrees of freedom for the sample variance described above are (n 1).
the imprecision may be negligible; (4) the bias may be negligible while
the imprecision is not negligible. within-laboratory standard deviation, nthe standard de-
viation of test results obtained within a laboratory for a
unit, nan object on which a measurement or observation single material under conditions that may include such
may be made. elements as different operators, equipment, and longer time
DISCUSSIONThe word unit is commonly used in the sense of a intervals.
unit of product (service, etc.)the entity of product inspected in order
NOTE 59Because the training of operators, the agreement of different
to determine its classification or its measurements. This entity may be
pieces of equipment in the same laboratory and the variation of environ-
a single article, a set of like articles treated collectively, a subassembly,
mental conditions with longer time intervals depend on the degree of
a stated quantity of material, etc. The unit of product or service need not
within-laboratory control, the within-laboratory standard deviation is
be the same as the unit of purchase, supply, production, or shipment.
likely to vary appreciably from laboratory to laboratory.
universe (population), nthe totality of the set of items, Youden square, na type of block design derived from
units, or measurements, etc., real or conceptual, that is under certain Latin squares by deleting, or adding, rows (or
consideration. columns) so that one block factor remains complete blocks
NOTE 56This definition of universe is being revised to incorporate the and the second block factor constitutes balanced incomplete
concept of including one or more populations. Use with caution. blocks. E 1325

Copyright ASTM International 9Document provided by IHS Licensee=Instituto Mexicano Del Petroleo/3139900100,
Reproduced by IHS under license with ASTM 10/27/2004 13:11:48 MDT Questions or comments about this message: please call
the Document Policy Group at 303-397-2295.
E 456 02e1
ANNEX

(Mandatory Information)

A1. MEASUREMENT TERMINOLOGY

A1.1 A test method often has three distinct stages: (1) the test method specifies that only one test determination is to be
direct observation of dimensions or characteristics, (2) the made, then the test determination value is the test result of the
combining of the observed values to obtain a single test test method. Some test methods require that several determi-
determination, and (3) the combining of a number of test nations be made and the values obtained be averaged or
determinations to obtain the test result of the test method. The otherwise combined to obtain the test result of the test method.
term measurement may be applied to any one or more of these Averaging of several determinations is often used to reduce the
stages of the measurement process. effect of local variations of the property within the material.
A1.2 In the simplest of test methods a single direct
observation is also the test determination and the test result. A1.3 Precision statements for ASTM test methods are
For example, a test observation required by a test method may usually based on test results, not test determinations or obser-
be the mass of a test specimen prepared and weighed in a vations. If for some compelling reason an ASTM committee
specified way. The observation would also be the test determi- wished to address the issue of variation between test determi-
nation of the mass of the test specimen, and if only one nations (in addition to the variation among test results), the
specimen is to be weighed, the observed weight would also be committee can do so with a clear declaration (of what is being
the test result of the test method. Another test method may done) to avoid confusion. Sampling plans and product speci-
require the measurement of the area of the test specimen as fications should specify the sample size in terms of the number
--`,,,````,,````,``,,```,,`,-`-`,,`,,`,`,,`---

well as the mass, and then direct that the mass be divided by of replicate test results. A test method should specify the
the area to obtain the mass per unit area of the test specimen. required observations to obtain a test determination and the
The whole process of measuring the mass and the area and number of test determinations to be averaged or otherwise
calculating the mass per unit area is a test determination. If the combined to obtain a single test result.

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 service@astm.org (e-mail); or through the ASTM website
(www.astm.org).

Copyright ASTM International 10Document provided by IHS Licensee=Instituto Mexicano Del Petroleo/3139900100,
Reproduced by IHS under license with ASTM 10/27/2004 13:11:48 MDT Questions or comments about this message: please call
the Document Policy Group at 303-397-2295.

You might also like