QA GlossaryThis is the Glossary of QA Terms of the Quality Assurance Management Staff (QAMS) within the Office of Modeling, Monitoring Systems, and Quality Assurance in the EPA's Office of Research and Development. To use this service, simply "click" on a letter below that corresponds to the first letter of the term you are searching for. This will bring you to the section for that letter.
Permission to publish provided by Fred Haeberer, QAMS, who maintains this QAMS Glossary as a working document. Questions or comments about these terms can be directed to Mr. Haeberer at 202-260-5785.
Absolute method: a body of procedures and techniques for which measurement is based entirely on physically defined, fundamental quantities. Acceptable quality level: a limit above which quality is considered satisfactory and below which it is not. In sampling inspection, the maximum percentage of defects or failures that can be considered satisfactory as an average. Acceptable quality range: the interval, between specified upper and lower limits of a sequence of values, within which the values are considered to be satisfactory. Acceptable value: an observed or corrected value that falls within the acceptable range. See Corrected value and Observed value. Acceptance sampling: the procedure of drawing samples from a lot or population to determine whether to accept or reject a sampled lot or population. Accepted reference value: a numerical quantity that serves as an agreed-upon basis for comparison, and which is derived as: l) a theoretical or established quantity based on scientific principles, 2) an assigned value, based on experimental work of some recognized organization, or 3) a consensus quantity based on collaborative experimental work under the auspices of a scientific or engineering group. Accreditation: a formal recognition that an organization (e.g., laboratory) is competent to carry out specific tasks or specific types of tests. See also Certification. Accreditation criterion: a requirement that a laboratory must meet to receive authorization and approval to perform a specified task. Accredited laboratory: a laboratory which has been evaluated and given approval to perform a specified measurement or task, usually for a specific property or analyte and for a specified period of time. Accuracy: the degree of agreement between an observed value and an accepted reference value. Accuracy includes a combination of random error (precision) and systematic error (bias) components which are due to sampling and analytical operations; a data quality indicator. EPA recommends that this term not be used and that precision and bias be used to convey the information usually associated with accuracy. See Precision and Bias. Action limit: See Control limit. Adjusted value: the observed value after adjustment for values of a blank or bias of the measurement system. Aliquant: a subsample derived by a divisor that divides a sample into a number of equal parts but leaves a remainder; a subsample resulting from such a divisor. See Subsample.
Aliquot: a subsample derived by a divisor that divides a sample into a number of equal parts and leaves no remainder; a subsample resulting from such a division. In analytical chemistry the term aliquot is generally used to define any representative portion of the sample.
Alpha error: See Type I Error.
Alternate method: any body of procedures and techniques of sample collection and/or analysis for a characteristic of interest which is not a reference or approved equivalent method but which has been demonstrated in specific cases to produce results comparable to those obtained from a reference method.
Analysis (chemical): the determination of the qualitative and/or quantitative composition of a substance.
Analyte: the substance, a property of which is to be measured by chemical analysis.
Analytical batch: a group of samples, including quality control samples, which are processed together using the same method, the same lots of reagents, and at the same time or in continuous, sequential time periods. Samples in each batch should be of similar composition and share common internal quality control standards.
Analytical blank: See Reagent blank.
Analytical limit of discrimination: See Method detection limit.
Analytical reagent (AR): the American Chemical Society's designation for the highest purity of certain chemical reagents and solvents. See Reagent grade.
Arithmetic mean: the sum of all the values of a set of measurements divided by the number of values in the set, usually denoted by x-; a measure of central tendency. See Measure of central tendency.
Assignable cause: a factor or an experimental variable shown to significantly change the quality of an effect or a result.
Audit: a systematic evaluation to determine the conformance to quantitative specifications of some operational function or activity. See Audit of data quality, Performance evaluation audit, and Technical systems audit, and also Review, and Management systems review.
Audit of data quality (ADQ): a qualitative and quantitative evaluation of the documentation and procedures associated with environmental measurements to verify that the resulting data are of acceptable quality.
Audit sample: See Performance evaluation sample.
Average: See Arithmetic mean.
Background level (environmental): the concentration of substance in a defined control area during a fixed period of time before, during or after a data gathering operation.
Batch: a quantity of material produced or processed in one operation, considered to be a uniform discrete unit.
Batch-lot: the samples collected under sufficiently uniform conditions to be processed as a group. See Batch, Batch size.
Batch-sample: one of the samples drawn from a batch.
Batch-size: the number of samples in a batch-lot.
Beta error: See Type II Error.
Bias: the systematic or persistent distortion of a measurement process which deprives the result of representativeness (i.e., the expected sample measurement is different than the sample's true value.) A data quality indicator.
Blank sample: a clean sample or a sample of matrix processed so as to measure artifacts in the measurement (sampling and analysis) process.
Blind sample: a subsample submitted for analysis with a composition and identity known to the submitter but unknown to the analyst and used to test the analyst's or laboratory's proficiency in the execution of the measurement process. See Double-blind sample.
Bulk sample: a sample taken from a larger quantity (lot) for analysis or recording purposes.
Calibrant: See Calibration standard.
Calibrate: to determine, by measurement or comparison with a standard, the correct value of each scale reading on a meter or other device, or the correct value for each setting of a control knob. The levels of the calibration standards should bracket the range of planned measurements. See Calibration curve.
Calibration-check: See Calibrate.
Calibration-check standard: See Calibration standard.
Calibration curve: the graphical relationship between the known values for a series of calibration standards and instrument responses.
Calibration drift: the difference between the instrument response and a reference value after a period of operation without recalibration.
Calibration standard: a substance or reference material used to calibrate an instrument.
Candidate method: a body of procedures and techniques of sample collection and/or analysis that is submitted for approval as a reference method, an equivalent method, or an alternative method.
Carrying-agent: any diluent or matrix used to entrain, dilute or to act as a vehicle for a compound of interest.
CAS#: Chemical Abstracts Service registry number of elements, chemical compounds, and certain mixtures.
Cause-effect diagram: a graphical representation of an effect and possible causes. A popular one is the Ishikawa "fish bone diagram."
Central line: the line on a control chart that represents the expected value of the control chart statistic; often the mean. See Control chart.
Certification: the process of testing and evaluation against specifications designed to document, verify, and recognize the competence of a person, organization, or other entity to perform a function or service usually for a specified time. See also Accreditation.
Certification of Data Quality: the real-time attestation that the activities of an environmental data collection operation's individual elements (e.g., sampling design, sampling, sample handling, chemical analysis, data reduction, etc.,) have been carried out in accordance with the operation's requirements and that the results meet the defined quality criteria.
Certified Reference Material (CRM): a reference material that has one or more of its property values established by a technically valid procedure and is accompanied by or traceable to a certificate or other documentation issued by a certifying body. See Certification and Reference material.
Certified value: the reported numerical quantity that appears on a certificate for a property of a reference material.
Chain-of-custody: an unbroken trail of accountability that insures the physical security of samples, data and records.
Chance cause: an unpredictable, random determinant of variation of a response in a sampling or measurement operation.
Characteristic: See Property.
Check sample: an uncontaminated sample matrix spiked with known amounts of analytes usually from the same source as the calibration standards. It is generally used to establish the stability of the analytical system but may also be used to assess the performance of all or a portion of the measurement system. See also Quality control sample.
Check standard: a substance or reference material obtained from a source independent from the source of the calibration standard; used to prepare check samples.
Chi-square test: a statistical test of the agreement between the observed frequency of events and the frequency expected according to some hypothesis.
Clean sample: a sample of a natural or synthetic matrix containing no detectable amount of the analyte of interest and no interfering material.
Coefficient of variation (CV): a measure of relative dispersion (precision.) It is equal to the ratio of the standard deviation divided by the arithmetic mean. See also Relative standard deviation.
Collaborative testing: the evaluation of an analytical method by typical or representative laboratories using subsamples prepared from a homogeneous standard sample.
Collocated sample: one of two or more independent samples collected so that each is equally representative for a given variable at a common space and time.
Collocated samplers: two or more identical sample collection devices, located together in space and operated simultaneously, to supply a series of duplicate or replicate samples for estimating precision of the total measurement system/process.
Comparability: the degree to which different methods, data sets and/or decisions agree or can be represented as similar; a data quality indicator.
Completeness: the amount of valid data obtained compared to the planned amount, and usually expressed as a percentage; a data quality indicator.
Component of variance: a part of the total variance associated with a specified source of variation.
Composite sample: a sample prepared by physically combining two or more samples having some specific relationship and processed to ensure homogeneity. See Flow-proportioned sample and Time-proportioned sample.
Confidence coefficient: the probability statement that accompanies a confidence interval and is equal to unity minus the associated type I error rate (false positive rate). A confidence coefficient of 0.90 implies that 90% of the intervals resulting from repeated sampling of a population will include the unknown (true) population parameter. See Confidence interval.
Confidence interval: the numerical interval constructed around a point estimate of a population parameter, combined with a probability statement (the confidence coefficient) linking it to the population's true parameter value. If the same confidence interval construction technique and assumptions are used to calculate future intervals, they will include the unknown population parameter with the same specified probability. See Confidence coefficient.
Control chart: a graph of some measurement plotted over time or sequence of sampling, together with control limit(s) and, usually, a central line and warning limit(s). See Central line, Control limit and Warning limit.
Control limit: a specified boundary on a control chart that, if exceeded, indicates a process is out of statistical control, and the process must be stopped, and corrective action taken before proceeding (e.g., for a Shewhart- chart the control limits are the mean plus and minus three standard deviations, i.e., the 99.72% confidence level on either side of the central line.)
Control sample: See Quality control sample and Check sample.
Control standard: See Check standard.
Controlled variable: a variable that is set at a pre-selected level when a controlled experiment is conducted.
Correlation: a measure of association between two variables. See also Correlation coefficient.
Correlation coefficient: a number between -1 and 1 that indicates the degree of linearity between two variables or sets of numbers. The closer to -1 or +1, the stronger the linear relationship between the two (i.e., the better the correlation.) Values close to zero suggest no correlation between the two variables. The most common correlation coefficient is the product-moment, a measure of the degree of linear relationship between two variables.
Critical-toxicity range: the interval between the highest concentration at which all test organisms survive and the lowest concentration at which all test organisms die within the test period.
Daily standard: synonym for Calibration standard.
Data: facts or figures from which conclusions can be inferred.
Data quality: the totality of features and characteristics of data that bears on their ability to satisfy a given purpose; the sum of the degrees of excellence for factors related to data.
Data quality indicators: quantitative statistics and qualitative descriptors that are used to interpret the degree of acceptability or utility of data to the user. The principal data quality indicators are bias, precision, accuracy, comparability, completeness, and representativeness.
Data Quality Objective (DQO): qualitative and quantitative statements of the overall level of uncertainty that a decision-maker is willing to accept in results or decisions derived from environmental data. DQOs provide the statistical framework for planning and managing environmental data operations consistent with the data user's needs.
Data reduction: the process of transforming raw data by arithmetic or statistical calculations, standard curves, concentration factors, etc., and collation into a more useful form.
Data set: all the observed values for the samples in a test or study; a group of data collected under similar conditions and which, therefore, can be analyzed as a whole.
Datum: the singular of data. See Data and Value.
Defensible: the ability to withstand any reasonable challenge related to the veracity or integrity of laboratory documents and derived data.
Degrees of freedom: the total number of items in a sample minus the number of independent relationships existing among them; the divisor used to calculate a variance term; in the simplest cases, it is one less than the number of observations.
Dependent variable: See Response variable.
Detection limit (DL): the lowest concentration or amount of the target analyte that can be determined to be different from zero by a single measurement at a stated level of probability. See Method detection limit.
Determination: the application of the complete analytical process of measuring the property of interest in a sample, from selecting or measuring a test portion to the reporting of results. See Test determination.
Diluent: a substance added to another to reduce the concentration and resulting in a homogeneous end product without chemically altering the compound of interest.
Dilution factor: the numerical value obtained from dividing the new volume of a diluted substance by its original volume.
Document control: a systematic procedure for indexing documents by number, date and revision number for archiving, storage, and retrieval.
Double-blind sample: a sample submitted to evaluate performance with concentration and identity unknown to the analyst. See Blind sample.
Duplicate: an adjective describing the taking of a second sample or performance of a second measurement or determination. Often incorrectly used as a noun and substituted for "duplicate sample." Replicate is to be used if there are more than two items. See Replicate.
Duplicate analyses or measurements: the analyses or measurements of the variable of interest performed identically on two subsamples of the same sample. The results from duplicate analyses are used to evaluate analytical or measurement precision but not the precision of sampling, preservation or storage internal to the laboratory.
Duplicate samples: two samples taken from and representative of the same population and carried through all steps of the sampling and analytical procedures in an identical manner. Duplicate samples are used to assess variance of the total method including sampling and analysis. See Collocated sample.
Dynamic blank: a sample-collection material or device (e.g., filter or reagent solution) that is not exposed to the material to be selectively captured but is transported and processed in the same manner as the sample. See Instrumental blank and Sampling equipment blank.
Dynamic calibration: standardization of both the measurement and collection systems using a reference material similar to the unknown. For example, a series of air-mixture standards containing sulfur dioxide of known concentrations could be used to calibrate a sulfur dioxide bubbler system.
Environmental sample: a sample of any material that is collected from an environmental source.
Environmentally related measurement: any assessment of environmental concern generated through or for field, laboratory, or modelling processes; the value obtained from such an assessment.
Equivalent method: any method of sampling and/or analysis demonstrated to result in data having a consistent and quantitatively known relationship to the results obtained with a reference method under specified conditions, and formally recognized by the EPA.
Error (measurement): the difference between an observed or corrected value of a variable and a specified, theoretically correct, or true value.
Error function: the mathematical relationship of the results obtained from the measurement of one or more properties and the error of the applied measurement process. See Normal distribution.
Experimental variable: See Independent variable.
External quality control: the activities which are routinely initiated and performed by persons outside of normal operations to assess the capability and performance of a measurement process.
False negative decision: See Type II Error.
False negative result: estimating (incorrectly) that an analyte is not present when it actually is present.
False positive decision: See Type I Error.
False positive result: estimating (incorrectly) that an analyte is present when it is actually not present.
Field blank: a clean sample (e.g., distilled water), carried to the sampling site, exposed to sampling conditions (e.g., bottle caps removed, preservatives added) and returned to the laboratory and treated as an environmental sample. Field blanks are used to check for analytical artifacts and/or background introduces by sampling and analytical procedures. See Dynamic blank and Sampling equipment blank.
Field duplicates: See Duplicate sample.
Field reagent blank: See Field blank.
Field sample: See Sample.
Flow rate: the quantity-per-unit time of a substance passing a point, plane, or space; for example the volume or mass of gas or liquid emerging from an orifice, pump, or turbine or moving through a point in a conduit or channel.
Flow-proportioned sample: a sample or subsample collected from a fluid system at a rate that produces a constant ratio of sample accumulation to matrix flow rate.
Fortify: synonym for Spike.
Full-scale response: the maximum output of a measurement instrument in a given range as displayed on a meter or scale.
Functional analysis: a mathematical evaluation of each component of the measurement system (sampling and analysis) in order to quantitate the error for each component. A functional analysis is usually performed prior to a ruggedness test in order to determine those variables which should be studied experimentally.
Geometric mean: the antilogarithm of the mean of the logarithms of all the values in a set.
Good laboratory practices (GLP): either general guidelines or formal regulations for performing basic laboratory operations or activities that are known or believed to influence the quality and integrity of the results.
Goodness-of-fit: the measure of agreement between the data in a data set and the expected or hypothesized values.
Grab sample: a single sample which is collected at one point in time and place.
Gross sample: See Bulk sample.
Homogeneity: the degree of uniformity of structure or composition.
In-control: a condition indicating that performance of the quality control system is within the specified control limits, i.e., that a stable system of chance is operating and resulting in statistical control. See Control chart.
Independent variable: See Controlled variable.
Inspection criterion: the specification(s) and rationale for rejecting and accepting samples in a particular sampling plan.
Instrument blank: a clean sample processed through the instrumental steps of the measurement process; used to determine instrument contamination. See Dynamic blank.
Interference: a positive or negative effect on a measurement caused by a variable other than the one being investigated.
Interference equivalent: the mass or concentration of a foreign substance which gives the same measurement response as one unit of mass or concentration of the substance being measured.
Interlaboratory calibration: the process, procedures, and activities for standardizing a given measurement system to ensure that laboratories participating in the same program can produce comparable data.
Interlaboratory method validation study (IMVS): the formal study of a sampling and/or analytical method, conducted with replicate, representative matrix samples, following a specific study protocol and utilizing a specific written method, by a minimum of seven laboratories, for the purpose of estimating interlaboratory precision, bias and analytical interferences.
Interlaboratory precision: a measure of the variation, usually given as the standard deviation, among the test results from independent laboratories participating in the same test.
Interlaboratory test: a test performed by two or more laboratories on the same material for the purpose of assessing the capabilities of an analytical method or for comparing different methods.
Internal quality control: See Intralaboratory quality control.
Internal standard: a standard added to a test portion of a sample in a known amount and carried through the entire determination procedure as a reference for calibration and controlling the precision and bias of the applied analytical method.
Intralaboratory quality control: the routine activities and checks, such as periodic calibrations, duplicate analyses and spiked samples, that are included in normal internal procedures to control the accuracy and precision of measurements.
Intralaboratory precision: a measure of the method/sample specific analytical variation within a laboratory; usually given as the standard deviation estimated from the results of duplicate/replicate analyses. See also Standard deviation and Variance.
Laboratory accreditation: See Accredited laboratory and Accreditation.
Laboratory blank: See Reagent blank.
Laboratory control sample: See Quality control sample.
Laboratory duplicates: synonym for Duplicate analyses.
Laboratory performance check solution: a solution of method and surrogate analytes and internal standards; used to evaluate the performance of the instrument system against defined performance criteria.
Laboratory replicates: See Replicate analysis or measurement.
Laboratory spiked blank: See Spiked laboratory blank.
Laboratory spiked sample: See Spiked sample.
Laboratory sample: a subsample of a field, bulk or batch sample selected for laboratory analysis.
Least squares method: a technique for estimating model coefficients which minimizes the sum of the squares of the differences between each observed value and its corresponding predicted value derived from the assumed model.
Limit of detection (LOD): See Method detection limit.
Limit of quantification (LOQ): the concentration of analyte in a specific matrix for which the probability of producing analytical values above the method detection limit is 99 percent.
Linearity: the degree of agreement between the calibration curve of a method and a straight line assumption.
Lot: a number of units of an article or a parcel of articles offered as one item; commonly, one of the units, such as a sample of a substance under study. See Batch.
Lot size: the number of units in a particular lot. See Batch lot and Batch size.
Lower control limit: See Control limit.
Lower warning limit: See Warning limit.
Management systems review (MSR): the qualitative assessment of a data collection operation and/or organization(s) to establish whether the prevailing quality management structure, practices, and procedures are adequate for ensuring that the type and quality of data needed and expected are obtained. See Review and Audit
Matrix: a specific type of medium (e.g., surface water, drinking water) in which the analyte of interest may be contained. See Medium.
Matrix spike: See Spiked sample.
Matrix spike duplicate sample analysis: See Matrix, Duplicate analysis and Spiked sample.
Maximum contaminant level: the highest permissible concentration of a pollutant that may be delivered to any receptor.
Maximum holding time: the length of time a sample can be kept under specified conditions without undergoing significant degradation of the analyte(s) or property of interest.
Mean: See Arithmetic mean.
Measurement range: the range over which the precision and/or recovery of a measurement method are regarded as acceptable. See Acceptable quality range.
Measurement standard: a standard added to the prepared test portion of a sample (e.g. to the concentrated extract or the digestate) as a reference for calibrating and controlling measurement or instrumental precision and bias.
Measure of central tendency: a statistic that describes the grouping of values in a data set around some common value (e.g., the median, arithmetic mean, or geometric mean.)
Measure of dispersion: a statistic that describes the variation of values in a data set around some common value. See Coefficient of variation, Range, Variance and Standard deviation.
Medium: a substance (e.g., air, water, soil) which serves as a carrier of the analytes of interest. See Matrix.
Medium blank: See Field blank and/or Laboratory blank.
Median: the middle value for an ordered set of n values; represented by the central value when n is odd or by the mean of the two most central values when n is even.
Method: a body of procedures and techniques for performing a task (e.g., sampling, characterization, quantification) systematically presented in the order in which they are to be executed.
Method blank: a clean sample processed simultaneously with and under the same conditions as samples containing an analyte of interest through all steps of the analytical procedure.
Method check sample: See Spiked laboratory blank.
Method detection limit (MDL): the minimum concentration of an analyte that, in a given matrix and with a specific method, has a 99% probability of being identified, qualitatively or quantitatively measured, and reported to be greater than zero. See Detection limit.
Method of least squares: See Least squares method.
Method performance study: See Interlaboratory method validation study.
Method quantification limit (MQL): See Limit of quantification and also Method detection limit.
Minimum detectable level: See Method detection limit.
Mode: the most frequent value or values in a data set.
Multipoint calibration: the determination of correct scale values by measuring or comparing instrument responses at a series of standardized analyte concentrations; used to define the range for generating quantitative data of acceptable quality.
Noise: the sum of random errors in the response of a measuring instrument.
Normal distribution: an idealized probability density function that approximates the distribution of many random variables associated with measurements of natural phenomena and takes the form of a symmetric "bellshaped curve."
Observation: a fact or occurrence that is recognized and recorded.
Observed value: the magnitude of a specific measurement; a variable; a unit of space, time or quantity; a datum. The observed value is that reported before correction for a blank value. See Corrected value.
Outlier: an observed value that appears to be discordant from the other observations in a sample. One of a set of observations that appears to be discordant from the others. The declaration of an outlier is dependent on the significance level of the applied identification test. See also Significance level.
Parameter: any quantity such as a mean or a standard deviation characterizing a population. Commonly misused for "variable", "characteristic" or "property."
Percentage standard deviation: synonym for Relative standard deviation.
Performance evaluation audit: a type of audit in which the quantitative data generated in a measurement system are obtained independently and compared with routinely obtained data to evaluate the proficiency of an analyst or laboratory.
Performance evaluation sample (PE sample): a sample, the composition of which is unknown to the analyst and is provided to test whether the analyst/laboratory can produce analytical results within specified performance limits. See Blind sample and Performance evaluation audit.
Population: all possible items or units which possess a variable of interest and from which samples may be drawn.
Precision: the degree to which a set of observations or measurements of the same property, usually obtained under similar conditions, conform to themselves; a data quality indicator. Precision is usually expressed as standard deviation, variance or range, in either absolute or relative terms. See also Standard deviation and Variance.
Preventative maintenance: an orderly program of activities designed to ensure against equipment failure.
Primary reference standard: See Primary standard.
Primary standard: a substance or device, with a property or value that is unquestionably accepted (within specified limits) in establishing the value of the same or related property of another substance or device.
Probability: a number between zero and one inclusive, reflecting the limiting proportion of the occurrence of an event in an increasingly large number of identical trials, each of which results in either the occurrence or nonoccurrence of the event.
Probability sampling: sampling in which: (a) every member of the population has a known probability of being included in the sample; (b) the sample is drawn by some method of random selection consistent with these probabilities; and (c) the known probabilities of inclusion are used in forming estimates from the sample. The probability of selection need not be equal for members of the population.
Procedure: a set of systematic instructions for performing an operation.
Proficiency testing: a systematic program in which one or more standardized samples is analyzed by one or more laboratories to determine the capability of each participant.
Property: a quality or trait belonging and peculiar to a thing; a response variable is a measure of a property. Synonym for Characteristic.
Protocol: a detailed written procedure for a field and/or laboratory operation (e.g., sampling, analysis) which must be strictly adhered to.
Quality: the sum of features and properties/characteristics of a product or service that bear on its ability to satisfy stated needs.
Quality assessment: the evaluation of environmental data to determine if they meet the quality criteria required for a specific application.
Quality assurance (QA): an integrated system of activities involving planning, quality control, quality assessment, reporting and quality improvement to ensure that a product or service meets defined standards of quality with a stated level of confidence.
Quality Assurance Narrative Statement: a description of the quality assurance and quality control activities to be followed for a research project.
Quality Assurance Objectives: the limits on bias, precision, comparability, completeness and representativeness defining the minimal acceptable levels of performance as determined by the data user's acceptable error bounds.
Quality Assurance Program Plan (QAPP): a formal document describing the management policies, objectives, principles, organizational authority, responsibilities, accountability, and implementation plan of an agency, organization or laboratory for ensuring quality in its products and utility to its users.
Quality Assurance Project Plan (QAPjP): a formal document describing the detailed quality control procedures by which the quality requirements defined for the data and decisions pertaining to a specific project are to be achieved.
Quality Circle: a small group of individuals from an organization or unit who have related interests and meet regularly to consider problems or other matters related to the quality of the product or process.
Quality control (QC): the overall system of technical activities whose purpose is to measure and control the quality of a product or service so that it meets the needs of users. The aim is to provide quality that is satisfactory, adequate, dependable, and economical.
Quality control chart: See Control chart.
Quality control check sample: See Calibration standard.
Quality control sample: an uncontaminated sample matrix spiked with known amounts of analytes from a source independent from the calibration standards. It is generally used to establish intralaboratory or analyst specific precision and bias or to assess the performance of all or a portion of the measurement system. See also Check sample.
Quantitation limits: the maximum or minimum levels or quantities of a target variable that can be quantified with the certainty required by the data user.
Random: lacking a definite plan, purpose or pattern; due to chance.
Random error: the deviation of an observed value from a true value, which behaves like a variable in that any particular value occurs as though chosen at random from a probability distribution of such errors. The distribution of random error is generally assumed to be normal.
Random sample or subsample: a subset of a population or a subset of a sample, selected according to the laws of chance with a randomization procedure.
Random variable: a quantity which may take any of the values of a specified set with a specified relative frequency or probability. It is defined by a set of possible values, and by an associated probability function giving the relative frequency of occurrence of each possible value.
Randomization: the arrangement of a set of objects in a random order, a set of treatments applied to a set of experimental units is said to be randomized when the treatment applied to any given unit is chosen at random from those available and not already allocated.
Randomness: a basic statistical concept and property implying an absence of a plan, purpose or pattern, or of any tendency to favor one outcome rather than another.
Range: the difference between the minimum and the maximum of a set of values.
Raw data: any original factual information from a measurement activity or study recorded in laboratory worksheets, records, memoranda, notes, or exact copies thereof and that are necessary for the reconstruction and evaluation of the report of the activity or study. Raw data may include photographs, microfilm or microfiche copies, computer printouts, magnetic media, including dictated observations, and recorded data from automated instruments. If exact copies of raw data have been prepared (e.g., tapes which have been transcribed verbatim, dated, and verified accurate by signature), the exact copy or exact transcript may be substituted.
Reagent blank: a sample consisting of reagent(s), without the target analyte or sample matrix, introduced into the analytical procedure at the appropriate point and carried through all subsequent steps to determine the contribution of the reagents and of the involved analytical steps to error in the observed value.
Reagent grade: the second highest purity designation for reagents which conform to the current specifications of the American Chemical Society Committee on Analytical Reagents.
Records system (or plan): a written, documented group of procedures describing required records, steps for producing them, storage conditions, retention period and circumstances for their destruction or other disposition.
Recovery efficiency: in an analytical method, the fraction or percentage of a target analyte extracted from a sample containing a known amount of the analyte.
Reference material: a material or substance, one or more properties of which are sufficiently well established to be used for the calibration of an apparatus, the assessment of a measurement method, or assigning values to materials.
Reference method: a sampling and/or measurement method which has been officially specified by an organization as meeting its data quality requirements.
Reference standard: See Calibration standard.
Relative standard deviation: the standard deviation expressed as a percentage of the mean recovery, i.e., the coefficient of variation multiplied by 100.
Reliability: the likelihood that an instrument or device will function under defined conditions for a specified period of time.
Repeatability: the degree of agreement between mutually independent test results produced by the same analyst using the same test method and equipment on random aliquots of the same sample within a short period of time.
Replicability: See Repeatability.
Replicate: an adjective or verb referring to the taking of more than one sample or to the performance of more than one analysis. Incorrectly used as a noun in place of replicate analysis. Replicate is to be used when referring to more than two items. See Duplicate.
Replicate analyses or measurements: the analyses or measurements of the variable of interest performed identically on two or more subsamples of the same sample within a short time interval. See Duplicate analyses or measurements.
Replicate samples: two or more samples representing the same population characteristic, time, and place, which are independently carried through all steps of the sampling and measurement process in an identical manner. Replicate samples are used to assess total (sampling and analysis) method variance. Often incorrectly used in place of the term "replicate analysis." See Duplicate samples and Replicate analysis.
Representative sample: a sample taken so as to reflect the variable(s) of interest in the population as accurately and precisely as specified. To ensure representativeness, the sample may be either completely random or stratified depending upon the conceptualized population and the sampling objective (i.e., upon the decision to be made.)
Representativeness: the degree to which data accurately and precisely represent the frequency distribution of a specific variable in the population; a data quality indicator.
Reproducibility: the extent to which a method, test or experiment yields the same or similar results when performed on subsamples of the same sample by different analysts or laboratories.
Response variable: a variable that is measured when a controlled experiment is conducted.
Result: the product of a calculation, test method, test or experiment. The result may be a value, data set, statistic, tested hypothesis or an estimated effect.
Review: the assessment of management/operational functions or activities to establish their conformance to qualitative specifications or requirements. See Management systems review and also, Audit.
Risk: the probability or likelihood of an adverse effect.
Risk (statistical): the expected loss due to the use of a given decision procedure.
Robustness: (in)sensitivity of a statistical test method to departures from underlying assumptions. See Ruggedness.
Rounded number: a number, reduced to a specified number of significant digits or decimal places using defined criteria.
Routine method: a defined plan of procedures and techniques used regularly to perform a specific task.
Ruggedness: the (in)sensitivity of an analytical test method to departures from specified analytical or environmental conditions. See Robustness.
Ruggedness testing: the carefully ordered testing of an analytical method while making slight variations in test conditions (as might be expected in routine use) to determine how such variations affect test results. If a variation affects the results significantly, the method restrictions are tightened to minimize this variability.
Sample: a part of a larger whole or a single item of a group; a finite part or subset of a statistical population. A sample serves to provide data or information concerning the properties of the whole group or population.
Sample data custody: See Chain-of-custody.
Sample variance (statistical): a measure of the dispersion of a set of values. The sum of the squares of the difference between the individual values of a set and the arithmetic mean of the set, divided by one less than the number of values in the set. (The square of the sample standard deviation.) See also Measure of dispersion.
Sampling: the process of obtaining a representative portion of the material of concern.
Sampling equipment blank: a clean sample that is collected in a sample container with the sample-collection device and returned to the laboratory as a sample. Sampling equipment blanks are used to check the cleanliness of sampling devices. See Dynamic blank.
Sampling error: the difference between an estimate of a population value and its true value. Sampling error is due to observing only a limited number of the total possible values and is distinguished from errors due to imperfect selection, bias in response, errors of observation, measurement or recording, etc. See also Probability sampling.
Scheduled maintenance: See Preventative maintenance.
Screening test: a quick test for coarsely assessing a variable of interest.
Secondary standard: a standard whose value is based upon comparison with a primary standard.
Selectivity (analytical chemistry): the capability of a method or instrument to respond to a target substance or constituent in the presence of nontarget substances.
Sensitivity: capability of method or instrument to discriminate between measurement responses representing different levels of a variable of interest.
Significance level: the magnitude of the acceptable probability of rejecting a true null hypothesis or of accepting a false null hypothesis; the difference between the hypothetical value and the sample result.
Significant digit: any of the digits 0 through 9, excepting leading zeros and some trailing zeros, which is used with its place value to denote a numerical quantity to a desired rounded number. See Rounded number.
Significant figure: See Significant digit.
Single operator precision: the degree of variation among the individual measurements of a series of determinations by the same analyst or operator, all other conditions being equal.
Site: the area within boundaries established for a defined activity.
Span-drift: the change in the output of a continuous monitoring instrument over a stated time period during which the instrument is not recalibrated.
Span-gas: a gas of known concentration which is used routinely to calibrate the output level of an analyzer. See Calibration check standard.
Specimen: See Sample.
Spike: a known mass of target analyte added to a blank sample or subsample; used to determine recovery efficiency or for other quality control purposes.
Spiked laboratory blank: See Spiked reagent blank.
Spiked reagent blank: a specified amount of reagent blank fortified with a known mass of the target analyte; usually used to determine the recovery efficiency of the method.
Spiked sample: a sample prepared by adding a known mass of target analyte to a specified amount of matrix sample for which an independent estimate of target analyte concentration is available. Spiked samples are used, for example, to determine the effect of the matrix on a method's recovery efficiency.
Spiked sample duplicate analysis: See Duplicate analysis and Spiked sample.
Split samples: two or more representative portions taken from a sample or subsample and analyzed by different analysts or laboratories. Split samples are used to replicate the measurement of the variable(s) of interest.
Standard (measurement): a substance or material with a property quantified with sufficient accuracy to permit its use to evaluate the same property in a similar substance or material. Standards are generally prepared by placing a reference material in a matrix. See Reference material.
Standard addition: the procedure of adding known increments of the analyte of interest to a sample to cause increases in detection response. The level of the analyte of interest present in the original sample is subsequently established by extrapolation of the plotted responses.
Standard curve: See Calibration curve.
Standard deviation: the most common measure of the dispersion or imprecision of observed values expressed as the positive square root of the variance. See Variance.
Standard material: See Standard (measurement), Reference material.
Standard method: an assemblage of techniques and procedures based on consensus or other criteria, often evaluated for its reliability by collaborative testing and receiving organizational approval.
Standard operating procedure (SOP): a written document which details the method of an operation, analysis or action whose techniques and procedures are thoroughly prescribed and which is accepted as the method for performing certain routine or repetitive tasks.
Standard reference material (SRM): a certified reference material produced by the U.S. National Institute of Standards and Technology and characterized for absolute content independent of analytical method.
Standard reference sample: See Secondary standard.
Standard solution: a solution containing a known concentration of analytes, prepared and verified by a prescribed method or procedure and used routinely in an analytical method.
Standardization: the process of establishing the quantitative relationship between a known mass of target material (e.g., concentration) and the response variable (e.g., the measurement system or instrument response.) See Calibration, Calibration curve and Multipoint calibration.
Statistic: an estimate of a population characteristic calculated from a data set (observed or corrected values), e.g., the mean or standard deviation.
Stratification: the division of a target population into subsets or strata which are internally more homogeneous with respect to the characteristic to be studied than the population as a whole.
Stratified sampling: the sampling of a population that has been stratified, part of the sample coming from each stratum. See Stratification.
Stock solution: a concentrated solution of analyte(s) or reagent(s) prepared and verified by prescribed procedure(s), and used for preparing working standards or standard solutions.
Subsample: a representative portion of a sample. A subsample may be taken from any laboratory or a field sample. See Aliquant, Aliquot, Split sample and Test portion.
Surrogate analyte: a pure substance with properties that mimic the analyte of interest. It is unlikely to be found in environmental samples and is added to them for quality control purposes.
Surveillance: the act of maintaining supervision of or vigilance over a well-specified portion of the environment so that detailed information is provided concerning the state of that portion.
Synthetic sample: a manufactured sample. See Quality control sample.
Systematic error: a consistent deviation in the results of sampling and/or analytical processes from the expected or known value. Such error is caused by human and methodological bias.
Systems audit: See Technical systems audit.
Systems error: See Total systems error.
Target: the chosen object of investigation for which qualitative and/or quantitative data or information is desired, e.g., the analyte of interest.
Technical systems audit: a thorough, systematic on-site, qualitative review of facilities, equipment, personnel, training, procedures, record keeping, data validation, data management, and reporting aspects of a total measurement system.
Technique: a principle and/or the procedure of its application for performing an operation.
Test: a procedure used to identify or characterize a substance or constituent. See Method.
Test data: See Data.
Test determination: See Determination.
Test method: See Method.
Test portion: a subsample of the proper amount for analysis and measurement of the property of interest. A test portion may be taken from the bulk sample directly, but often preliminary operations, such as mixing or further reduction in particle size, are necessary. See Subsample.
Test result: a product obtained from performing a test determination. See Test determination.
Test sample: See Test portion.
Test specimen: See Test portion.
Test unit: See Test portion.
Time-proportioned sample: a composite sample produced by combining samples of a specific size, collected at preselected, uniform time intervals.
Total Quality Management (TQM): the process whereby an entire organization, led by senior management, commits to focusing on quality as a first priority in every activity. TQM implementation creates a culture in which everyone in the organization shares the responsibility for continuously improving the quality of products and services, (i.e., for "doing the right thing, the right way, the first time, on time.") in order to satisfy the customer.
Total measurement error: the sum of all the errors that occur from the taking of the sample through the reporting of results; the difference between the reported result and the true value of the population that was to have been sampled.
Traceability: an unbroken trail of accountability for verifying or validating the chain-of-custody of samples, data, the documentation of a procedure, or the values of a standard.
Treatment (experimental): an experimental procedure whose effect is to be measured and compared with the effect of other treatments.
Trip blank: a clean sample of matrix that is carried to the sampling site and transported to the laboratory for analysis without having been exposed to sampling procedures.
Tuning: the process of adjusting a measurement device or instrument, prior to its use, to ensure that it works properly and meets established performance criteria.
Type I error, (alpha error): an (incorrect) decision resulting from the rejection of a true hypothesis. (A false positive decision.)
Type II error, (beta error): an (incorrect) decision resulting from acceptance of a false hypothesis. (A false negative decision.)
Uncertainty: a measure of the total variability associated with sampling and measuring that includes the two major error components: systematic error (bias) and random error.
Universe: See Population.
Upper control limit: See Control limit.
Upper warning limit: See Warning limit.
User check: an evaluation of a written procedure (e.g., chemical analysis method) for clarity and accuracy in which an independent laboratory analyzes a small number of spiked samples, following the procedure exactly.
Valid study: a study conducted in accordance with accepted scientific methodology, the results of which satisfy predefined criteria.
Validated method: a method which has been determined to meet certain performance criteria for sampling and/or measurement operations.
Validation: the process of substantiating specified performance criteria.
Value: the magnitude of a quantity. A single piece of factual information obtained by observation or measurement and used as a basis of calculation.
Variable: an entity subject to variation or change.
Variance: See Sample variance.
Verifiable: the ability to be proven or substantiated.
Warning limit: a specified boundary on a control chart that indicates a process may be going out of statistical control and that certain precautions are required. For example, for a Shewhart- chart the warning limits are placed at plus and minus two standard deviations of the mean (i.e., at the 95% confidence interval.)
Working standard: See Secondary standard.
Zero drift: the change in instrument output over a stated time period of nonrecalibrated, continuous operation, when the initial input concentration is zero; usually expressed as a percentage of the full scale response.