1、Standard Practice for Definition of Terms Related to Quality and Statistics as Used in Highway Construction AASHTO Designation: R 10-06 (2016) Release: Group 1 (April 2016) American Association of State Highway and Transportation Officials 444 North Capitol Street N.W., Suite 249 Washington, D.C. 20
2、001 TS-5c R 10-1 AASHTO Standard Practice for Definition of Terms Related to Quality and Statistics as Used in Highway Construction AASHTO Designation: R 10-06 (2016) Release: Group 1 (April 2016) 1. SCOPE 1.1. This standard practice is to provide standard definitions for terms used in specification
3、s, test methods, and the various quality assurance procedures. 2. REFERENCED DOCUMENTS 2.1. ASTM Standards: E177, Standard Practice for Use of the Terms Precision and Bias in ASTM Test Methods E691, Standard Practice for Conducting an Interlaboratory Study to Determine the Precision of a Test Method
4、 E1325, Standard Terminology Relating to Design of Experiments 3. GENERAL PROVISIONS 3.1. It is recommended that AASHTO specifications and procedures, including standard practices, incorporate the following accepted terms as defined. These terms and their definitions are directed primarily to the va
5、rious quality assurance sampling and acceptance plans. Where a definition conflicts with a standard, the standard will govern. 4. TERMINOLOGY 4.1. acceptable quality level (AQL)the level of established actual quality for a quality characteristic that is fully acceptable. For example, when the qualit
6、y measure used is percent within limits (PWL), the AQL is the established (not estimated) PWL at which the quality characteristic is fully acceptable. Acceptance plans should be designed so that AQL material will receive an expected pay (EP) of 100 percent. 4.2. acceptancethe process whereby all fac
7、tors used by the agency (i.e., sampling, testing, and inspection) are evaluated to determine the degree of compliance with contract requirements and to determine the corresponding value for a given product. 4.3. acceptance constant (k)the minimum allowable quality index (Q). The acceptance constant
8、k is the acceptance limit associated with the quality index quality measure. In other words, for acceptance, Q must be greater than or equal to k. 4.4. acceptance limitin variables acceptance plans, the limiting upper or lower value, placed on a quality measure, that will permit acceptance of a lot.
9、 While the test values for material samples 2016 by the American Association of State Highway and Transportation Officials. All rights reserved. Duplication is a violation of applicable law.TS-5c R 10-2 AASHTO are evaluated against specification limits and engineering limits, the computed quality le
10、vel for a specific quality measure is evaluated against the acceptance limit. For example, the minimum allowable PWL called for in a QA specification might be 65 PWL. This is the acceptance limit. 4.5. acceptance number (c)in attributes acceptance plans, the maximum number of defective or nonconform
11、ing units in the sample that will permit acceptance of the inspected lot or batch. 4.6. acceptance planalso called “acceptance sampling plan” or “statistical acceptance plan. ” An agreed-upon process for evaluating the acceptability of a lot of material. It includes: lot size and sample size (i.e.,
12、number of samples), quality measure, acceptance limit(s), evaluation of risks, and pay adjustment provisions. 4.7. acceptance sampling and testingsampling and testing performed by the agency, or its designated agent, to evaluate acceptability of the final product. Also called “verification sampling
13、and testing” when specifically used to validate the contractors data. 4.8. acceptance quality characteristic (AQC)a quality characteristic that is measured and used to determine acceptability. 4.9. accepted reference valuea value that serves as an agreed-upon reference for comparison, and which is d
14、erived as: (1) a theoretical or established value, based on scientific principles; (2) an assigned or certified value, based on experimental work of some national or international organization; or (3) a consensus or certified value, based on collaborative experimental work under the auspices of a sc
15、ientific or engineering group. 4.10. accredited laboratorieslaboratories that are recognized by a formal accrediting body as meeting quality system requirements including demonstrated competence to perform standard test procedures. 4.11. action limitssee control chart limits. 4.12. accuracythe degre
16、e to which a measurement, or the mean of a distribution of measurements, tends to coincide with the true population mean. When the true population mean is not known, the degree of agreement between the observed measurements and an accepted reference value may be used to quantify the accuracy of the
17、measurements. 4.13. adjusted pay schedulesee pay adjustment schedule. 4.14. adjusted pay systemsee pay adjustment system. 4.15. aesthetic valuethe worth of an item of construction in terms of pleasing appearance or gratifying performance. 4.16. alpha () errorsee sellers risk. 4.17. alpha sub PF (PF)
18、see sellers risk. For acceptance plans with pay adjustments, the probability that acceptable quality level (AQL) material or construction will be assessed a pay factor less than PF with respect to a single acceptance quality characteristic. For example, 100for a compressive strength pay adjustment a
19、cceptance plan is the probability that AQL compressive strength material will be assessed less than 100 percent pay for compressive strength; this probability includes the probability of rejection, 0, with respect to compressive strength. The use of PFaddresses the need to quantify, for pay adjustme
20、nt acceptance plans, the probabilities associated with each possible pay decision regarding AQL material (e.g., payment of 0, 70, 80, 90, 100, or 105 percent for AQL material). A seller (contractor) is usually most interested in 100. 2016 by the American Association of State Highway and Transportati
21、on Officials. All rights reserved. Duplication is a violation of applicable law.TS-5c R 10-3 AASHTO 4.18. alternative hypothesis (Ha)the hypothesis to be accepted if the null hypothesis (Ho) is disproved (i.e., rejected). 4.19. analysis of variance (ANOVA)a mathematical method of isolating causes of
22、 variation. 4.20. assignable causea source of variation, usually due to error or process change, which can be detected by statistical methods and corrected within economic limits. When assignable causes are identified and removed, the production process is “under control.” 4.21. attributea character
23、istic that, by its presence or absence, classifies a unit or segment as conforming or nonconforming. 4.22. attributes acceptance plana statistical acceptance procedure where the acceptability of a lot of material or construction is evaluated by (1) noting the presence or absence of some characterist
24、ic or attribute in each of the units or samples in the group under consideration, and (2) counting how many units do or do not possess this characteristic. 4.23. averagea measure of central value that usually refers to the arithmetic mean ( X ). May also refer to the median or mode. The mean, median
25、, and mode are equal for a normal distribution. As a distribution becomes more skewed, the mean, median, and mode will differ more and more. 4.24. average absolute deviation (AAD)for a series of test results, the mean of absolute deviations from a target or specified value. Its use as an acceptance
26、measure is not recommended. 4.25. average run length (ARL)in a sample sense, the average number of times that a process will have been sampled and evaluated before a shift in process level is signaled. In a unit sense, the average number of units that will have been produced before a shift in level
27、is signaled. A long ARL is desirable for a process located at its specified level (to minimize calling for unneeded investigation or corrective action) and a short ARL is desirable for a process shifted to some undesirable level (so that corrective action will be called for promptly). ARL curves are
28、 used to describe the relative quickness in detecting level shifts of various control chart systems. 4.26. batcha specified quantity of material produced through a uniform operation or process. A batch is usually smaller than a lot. 4.27. beta () errorsee buyers risk. 4.28. beta sub PF (PF)see buyer
29、s risk. For acceptance plans with pay adjustments, the probability that rejectable quality level (RQL) material will be assessed a pay factor greater than or equal to PF. For example, 100for a compressive strength pay adjustment acceptance plan is the probability that RQL compressive strength materi
30、al will be assessed 100 percent pay or more for compressive strength. The use of PFaddresses the need to quantify, for pay adjustment acceptance plans, the probabilities associated with each possible pay decision regarding RQL material (e.g., payment of 0, 70, 80, 90, 100, or 105 percent for RQL mat
31、erial). A buyer (agency) is usually most interested in 100. 4.29. biasan error, constant in direction, that causes a measurement, or the mean of a distribution of measurements, to be offset from the true population mean. 4.30. biased samplinga sampling procedure whereby samples obtained from a lot d
32、o not have an equal probability of being chosen. Some examples of conventional biased approaches to sampling that have been used by agencies and contractors include representative sampling, uniform interval sampling, quota sampling, and selective sampling. 2016 by the American Association of State H
33、ighway and Transportation Officials. All rights reserved. Duplication is a violation of applicable law.TS-5c R 10-4 AASHTO 4.31. buyers risk ()also called “agencys risk or risk of a Type II” or “beta () error.” It is the risk to the agency of accepting rejectable quality level (RQL) material or work
34、manship. For an accept/reject acceptance plan, it is the probability that an acceptance plan will erroneously accept RQL material or workmanship with respect to a single acceptance quality characteristic. For variables acceptance plans using adjusted pay schedules, it is equivalent to PF, where PF =
35、 100. It is the probability that a variable payment acceptance plan will erroneously fully accept RQL material or workmanship at 100 percent pay or greater with respect to a single acceptance quality characteristic. 4.32. calibrationa set of operations that establish, under specified conditions, the
36、 relationship between values of quantities indicated by a measuring instrument or measuring system, or between values represented by a material measure or a reference material, and the corresponding values realized by standards. The purpose of calibration is to ensure that measurements made by the l
37、aboratory are traceable to the International System of Units (SI). Where traceability of measurements to SI units is not possible or relevant, measurements must be traceable to certified reference materials, agreed-upon methods, or consensus standards. Uncertainty estimates obtained during calibrati
38、on are used to judge if an instrument is suitable for its intended purpose. There is a need to re-establish traceability or recalibrate only when instrument measurements drift out of control (as determined through verification of calibration). See standardization. 4.33. certified techniciana technic
39、ian certified by some agency as proficient in performing certain duties. A certified technician is considered to be qualified. A qualified technician may or may not be certified. See qualified technician. 4.34. chance causea source of variation that is inherent in any production process and cannot b
40、e eliminated because it is due to random, expected causes. 4.35. characteristica measurable property of a material, product, or item of construction. 4.36. classa group of data values (measurements, observations, or test values) that all satisfy a given set of conditions associated with a frequency
41、distribution. 4.37. class frequencythe number of data values falling into a particular class. 4.38. classificationthe acts of evaluating if an attribute or variable does or does not conform to requirements, and the documentation of results. 4.39. class intervalthe difference in value between the low
42、er and the upper limits of a given class. 4.40. class limitsthe upper and lower values that define a given class. 4.41. class midpointthe central value of a particular class. 4.42. cluster samplingthe random selection of clusters (i.e., groups or bundles) from a population. For each cluster selected
43、, random sampling of the contained elements may then be performed. Examples of cluster sampling are selection of city blocks as primary sampling units; selection of a household as a cluster of people (of which only one may be interviewed); selection of bundles of rods or pipe from a shipment; and se
44、lection, from a shipment, of cartons that contain boxes or packages within them. 4.43. coefficient of determination (r2)a measure of the relationship between a single dependent random variable or response Y and a known independent variable X. It represents the proportion of the total variation of Y
45、due to X. For instance, if r2 = 0.81 (r = 0.9. See correlation coefficient), then 81 percent of the variation in the values of Y may be accounted for by the relationship with the variable X. The value of r2 from a regression model cannot be evaluated as “good” or “bad” 2016 by the American Associati
46、on of State Highway and Transportation Officials. All rights reserved. Duplication is a violation of applicable law.TS-5c R 10-5 AASHTO in singularity; it can only be judged relative to other models that have been estimated on similar phenomena. Thus, an r2of 0.30 for one phenomenon might be extreme
47、ly informative, while for another phenomenon it might be uninformative. 4.44. coefficient of variation (v)the ratio of the standard deviation to the mean expressed as a percent. It provides a measure of dispersion or spread relative to the mean. 4.45. combined pay factorsee composite pay factor. 4.4
48、6. component of variancea part of the total variance identified with a specified source of variability. 4.47. composite pay factoralso called “combined pay factor or overall pay factor.” A multiplication factor, often expressed as a percentage, that considers two or more quality characteristics and
49、is used to determine the contractors final payment for a unit of work. It is usually established using a performance model and life-cycle cost analysis. 4.48. computer simulationuse of a computer to generate conditions approximating actual or operational conditions. Computer simulation is a powerful and convenient tool to solve certain problems that are intractable by other methods. 4.49. confidence intervalan estimate of an interval in which the estimated parameter will
