What is TUR in calibration?

TUR, the test uncertainty ratio, compares the tolerance of the item being calibrated with the expanded uncertainty of the calibration at about 95 percent coverage. For a tolerance of ±T it is T / U, as ILAC-G8 writes it. A tolerance of ±0.010 mm checked with an expanded uncertainty of 0.002 mm gives a TUR of 5:1. A low TUR calls for a guard band or a better calibration process.

What is the difference between calibration and verification?

Calibration measures how far an instrument reads from a reference and states the uncertainty of that result. Verification takes the result and decides whether the instrument meets a stated requirement, such as your tolerance. The VIM defines them as separate operations and warns against mixing them up. A calibration certificate can report results with no pass or fail at all.

What do as-found and as-left mean?

As-found is the reading an instrument gives when it arrives for calibration, before anyone adjusts it. As-left is the reading after the work is done. As-found tells you whether the instrument was measuring correctly during the interval that just ended. As-left shows it is fit to go back into service. If nothing was adjusted, the two are the same.

What is a guard band in calibration?

A guard band is the gap between a tolerance limit and a tighter acceptance limit that the pass or fail decision is made against. It is sized from the measurement uncertainty, so an instrument passes only when its reading sits far enough inside the tolerance. JCGM 106 defines the term. ILAC-G8 notes that a guard band equal to the expanded uncertainty is common.

What is the difference between accuracy and precision?

In the VIM, accuracy is how close a measured value is to the true value, and precision is how closely repeated readings agree with each other. A gage can be precise and still inaccurate if it repeats the same offset. On data sheets and in the shop, accuracy usually means a stated limit such as ±0.02 mm, and precision is often used loosely for accuracy or for the number of digits on a display.

What does NIST traceable mean?

It means the results on a certificate can be related to standards maintained by NIST through a documented, unbroken chain of calibrations, each with a stated uncertainty. The claim is about the results, not the instrument. NIST does not certify or approve traceability claims made by others. The calibration provider has to support the claim and you have to assess it.

Compliance and audits

Calibration and metrology glossary

14 min read Published November 3, 2026

TUR (test uncertainty ratio) compares a tolerance with the uncertainty of the calibration that checks it. A guard band pulls the acceptance limits inside the tolerance to allow for that uncertainty, and as-found is the reading an instrument gives before anyone adjusts it. This glossary covers those and the other terms you meet on certificates and in audits, in plain words that follow the VIM and the GUM.

On this page

Calibration and its results

These terms describe the calibration event and the record it leaves behind.

Terms for calibration and its results
Term	Meaning
Calibration	Comparing an instrument with a reference standard of known value and uncertainty, and recording how far the instrument reads from it and how certain that result is. The VIM (JCGM 200:2012, entry 2.39) keeps adjustment out of the definition. A calibration on its own does not pass or fail anything.
Verification	Showing, with evidence, that an item meets the requirements set for it (VIM 2.44). In a calibration program it is the pass or fail judgment, made by comparing the calibration result with the tolerance the instrument has to hold. The VIM warns against confusing it with calibration.
Adjustment	Work on an instrument so that it reads correctly, such as setting zero, offset or span (VIM 3.11). It is not calibration. The VIM treats calibration as a prerequisite for adjustment and notes that an adjusted instrument usually has to be calibrated again.
As-found	The readings an instrument gives when it arrives for calibration, before any adjustment or repair. They show whether it was measuring correctly during the interval that just ended. ISO/IEC 17025:2017 clause 7.8.4.1 asks a calibration certificate to give results before and after any adjustment or repair where they are available.
As-left	The readings taken when the calibration work is finished, including any adjustment. They show the instrument is fit to go back into service and become the baseline for the next interval. When nothing was adjusted, as-left equals as-found.
Out-of-tolerance (OOT)	A calibration result outside the tolerance or maximum permissible error the instrument has to meet. An as-found OOT puts in doubt the measurements made since the last good calibration. ISO 9001 then asks you to determine whether earlier results were affected and to act on what you find.
Calibration certificate	The document that reports a calibration. ISO/IEC 17025:2017 clause 7.8.4.1 adds the measurement uncertainty, the conditions that affect the results, a statement of how the results are traceable, results before and after any adjustment or repair, and any statement of conformity. Clause 7.8.4.3 says it must not recommend a calibration interval unless that was agreed with the customer.
Calibration interval	The time, or the amount of use, allowed between two calibrations of an instrument. Quality standards ask for specified intervals but leave their length to you. ILAC-G24:2022 / OIML D 10:2022 lists the factors to weigh, among them drift, use, environment and calibration history, and describes methods for reviewing intervals from your own results.

Accuracy, precision and instrument behavior

The VIM gives these words narrower meanings than everyday speech. The difference matters when you read a data sheet against a certificate.

Terms for accuracy, precision and instrument behavior
Term	Meaning
Accuracy	How close a measured value is to the true value. In the VIM (2.13) accuracy is a quality and not a number: a measurement is more accurate when its error is smaller. On data sheets and in the shop, accuracy usually means a stated limit such as ±0.02 mm, which the VIM would call a maximum permissible error.
Precision	How closely repeated readings agree with each other under stated conditions (VIM 2.15). It is usually expressed as a standard deviation and says nothing about whether the readings are right. In shop talk precision often stands in for accuracy, a misuse the VIM points out, or for the number of digits on a display, which is resolution.
Bias	The difference between the average of repeated readings and a reference value. The VIM (2.18) calls bias an estimate of systematic error. In a gage study (AIAG MSA) it is measured on a reference part or master, and a known bias can be corrected.
Resolution	The finest change in the measured quantity that an instrument's reading can show (VIM 4.14). On a digital display it is the step of the last digit. The GUM (F.2.2.1) gives its standard uncertainty as about 0.29 times the resolution, and the term stays in a budget even when repeated readings are identical.
Stability	How well an instrument holds its metrological properties steady as time passes (VIM 4.19). In a gage study (AIAG MSA) it means how much the bias moves over time, usually tracked by measuring the same master at intervals on a control chart.
Drift	A gradual or stepwise change in an instrument's readings over time that comes from the instrument itself, not from the quantity measured or the conditions around it (VIM 4.21). The gap between one as-left and the next as-found shows it in practice. A drift rate is a direct input to setting a calibration interval.

Measurement uncertainty

The GUM (JCGM 100:2008) is the reference for evaluating and reporting uncertainty. These terms follow it in roughly the order a budget is built.

Terms for measurement uncertainty
Term	Meaning
Measurement uncertainty	A non-negative value that describes the spread of values that could reasonably be attributed to what you measured, based on the information available (VIM 2.26). It is not error. Error is the difference from a reference value, while uncertainty states how much doubt surrounds the result.
Standard uncertainty	An uncertainty expressed as one standard deviation (GUM 2.3.1, VIM 2.30). Each input to an uncertainty budget is converted to a standard uncertainty before the inputs are combined.
Type A evaluation	Evaluating an uncertainty by statistics on a series of readings, such as their standard deviation (GUM 2.3.2). When the result is the mean of n readings, its standard uncertainty is the standard deviation s divided by √n (GUM 4.2.3).
Type B evaluation	Evaluating an uncertainty by any other means: a calibration certificate, a manufacturer's specification, a handbook value, the resolution or experience (GUM 2.3.3). The split is about method only. The GUM (3.3.4) says it implies no difference in the nature of the components.
Sensitivity coefficient	The factor c that converts a change in one input into a change in the result, so each contribution becomes c × u. In the GUM (5.1.3) it is the partial derivative of the measurement model with respect to that input. It is how a temperature in °C turns into a length in µm.
Combined standard uncertainty	The standard uncertainty of the result, found by combining the standard uncertainties of all inputs, each multiplied by its sensitivity coefficient (GUM 2.3.4, VIM 2.31). For independent inputs it is the square root of the sum of the squared contributions (GUM 5.1.2).
Effective degrees of freedom (Welch-Satterthwaite)	A measure of how reliable a combined standard uncertainty is when some inputs come from only a few readings. The GUM (G.4.1) estimates it with the Welch-Satterthwaite formula, ν_eff = u_c⁴ / Σ(u_i⁴ / ν_i). The lower the result, the larger the coverage factor has to be.
Coverage factor k	The multiplier that turns the combined standard uncertainty into the expanded uncertainty (GUM 2.3.6, VIM 2.38). For a normal distribution k = 2 gives about 95 percent coverage, 95.45 percent in GUM Table G.1. With few effective degrees of freedom, k is taken from the t-distribution and is larger than 2.
Expanded uncertainty	The uncertainty you get by multiplying the combined standard uncertainty by a coverage factor, U = k × u_c (VIM 2.35). The result ± U gives a range that should hold most of the values the measured quantity could reasonably have (GUM 2.3.5). ILAC-P14 asks an accredited certificate to state it with its coverage factor and coverage probability.
Uncertainty budget	A table of the sources of uncertainty in one measurement that shows how each was evaluated, converted and combined into the stated uncertainty (VIM 2.33). It is the usual way to show an assessor where an uncertainty came from.

Tolerances, limits and decisions

These terms decide whether a reading counts as a pass. JCGM 106:2012 and ILAC-G8 are the main references.

Terms for tolerances, limits and decisions
Term	Meaning
Tolerance	The permitted range for a value, set by a drawing, a specification or your own requirement. Its ends are the tolerance limits. JCGM 106 (3.3.6) defines tolerance as the full difference between the upper and lower limits, while shops usually quote the half-width, as in ±0.01 mm.
MPE (maximum permissible error)	The largest error of indication that a specification or regulation allows for an instrument (VIM 4.26). Instrument standards and data sheets often state limits this way. The VIM says the word tolerance should not be used for an MPE, though shops often do.
TUR (test uncertainty ratio)	The tolerance of the item being calibrated divided by the expanded uncertainty of the calibration at about 95 percent coverage. ILAC-G8 writes it as the tolerance limit divided by U, so a tolerance of ±T gives TUR = T / U. NIST notes that MIL-STD-45662A adopted a 4:1 TUR in 1988 and ANSI/NCSL Z540-1 took it up in 1994.
TAR (test accuracy ratio)	An older ratio: the tolerance of the item divided by the stated accuracy of the reference standard alone. It leaves out resolution, repeatability, environment and method, so it can make a calibration look better than it is. TAR is shop usage and not a VIM or GUM term, so check which ratio a procedure means.
Decision rule	A lab's written rule for allowing for measurement uncertainty when it declares a result in or out of specification (ISO/IEC 17025:2017, 3.7). When a customer asks for a pass or fail, the lab has to define the rule and agree it with the customer unless the specification already fixes it (7.1.3). JCGM 106 and ILAC-G8 describe the common rules.
Statement of conformity	A pass or fail, or in or out of tolerance, statement against a specification. Under ISO/IEC 17025:2017 clause 7.8.6 the lab documents the decision rule behind it and the risk that rule carries. The report then names the results and the specification the statement covers and the rule applied.
Guard band	The gap you leave between a tolerance limit and the acceptance limit the pass or fail decision is made against (JCGM 106, 3.3.11). It is usually sized from the expanded uncertainty U. ILAC-G8 calls a guard band of w = U common and lists it as keeping the specific false accept risk below 2.5 percent.
False accept and false reject risk	A false accept passes an item that is really out of tolerance, and a false reject fails one that really conforms. JCGM 106 calls them consumer's and producer's risk, each either specific to one item or global across many. Under simple acceptance, ILAC-G8 notes, a result exactly on the tolerance limit can carry up to 50 percent false accept risk.

Gage studies and MSA

Measurement systems analysis (MSA) shows how much of your measured variation comes from the measurement system itself. These terms follow the AIAG Measurement Systems Analysis reference manual.

Terms for gage studies and MSA
Term	Meaning
Repeatability	The spread of readings when the same operator measures the same part with the same gage several times under the same conditions (VIM 2.20 and 2.21). In a Gage R&R (AIAG MSA) it is called equipment variation, EV.
Reproducibility	The spread between results when conditions change, such as a different operator, location or instrument (VIM 2.24 and 2.25). In a Gage R&R it is the variation between appraisers, called appraiser variation, AV.
GRR (gage R&R) and %GRR	GRR combines repeatability and reproducibility into one estimate of the measurement system's variation. %GRR expresses it as a percentage of the total variation in the study or of the tolerance, and the two bases give different results. The AIAG MSA manual gives acceptance guidelines for it.
ndc (number of distinct categories)	An estimate from a Gage R&R of how many separate groups of parts the measurement system can tell apart within the part variation in the study. It is calculated from the ratio of part variation to GRR. A low ndc means the gage cannot resolve real differences between parts.
Linearity	How an instrument's bias changes across its operating range (AIAG MSA). It is found by measuring reference parts at several points across the range and fitting a line to the bias at each point.

Standards, traceability and accreditation

These terms describe where a result gets its credibility: the chain of standards behind it and the competence of the lab that produced it.

Terms for standards, traceability and accreditation
Term	Meaning
Metrological traceability	The property of a measurement result that lets it be related to a reference, normally the SI unit, through a documented unbroken chain of calibrations, each adding to the measurement uncertainty (VIM 2.41). It belongs to a result. An instrument or a lab is not traceable in itself.
NIST traceable	Shorthand for results traceable through such a chain to standards maintained by NIST, the US national metrology institute. NIST does not certify or approve traceability claims made by others. The provider of a result has to back the claim and the user has to assess it.
Reference standard	The standard an organization or site keeps for calibrating its other standards of the same kind (VIM 5.6). In a shop it is often the top standard on site, calibrated by an outside laboratory.
Working standard	The standard you use day to day to calibrate or check instruments (VIM 5.7). It is usually calibrated against a reference standard. Gauge blocks used to check micrometers on the floor are a typical example.
CMC (calibration and measurement capability)	The calibration capability a lab offers customers under normal conditions, as described in its scope of accreditation or, for a national metrology institute, published in the BIPM key comparison database (ILAC-P14:09/2020). An accredited lab may not report a smaller uncertainty than its CMC. The uncertainty on your certificate tends to be larger because it includes your instrument's contribution.
Accreditation (ISO/IEC 17025)	Third-party attestation that a body is competent to carry out specific conformity assessment tasks (ISO/IEC 17011:2017, 3.1). Calibration labs are accredited to ISO/IEC 17025:2017, and the accreditation covers only the calibrations listed on the lab's scope. Certification to ISO 9001 is a different thing: it attests a management system, not technical competence.

Where Axiospec fits

Axiospec is calibration management software. It records each calibration with its as-found and as-left readings, the certificate and the due date, and it can keep an uncertainty budget for each instrument. It does not decide conformance or certify traceability. Those judgments stay with you, your customer and your assessor.

Sources

JCGM 200:2012, International vocabulary of metrology (VIM), third edition, published free by the BIPM. Entry numbers marked VIM refer to it.
JCGM 100:2008, Guide to the expression of uncertainty in measurement (GUM). Clause numbers marked GUM refer to it.
JCGM 106:2012, Evaluation of measurement data: the role of measurement uncertainty in conformity assessment.
ILAC-G8:09/2019, Guidelines on decision rules and statements of conformity.
ILAC-P14:09/2020, ILAC policy for measurement uncertainty in calibration.
ILAC-G24:2022 / OIML D 10:2022, Guidelines for the determination of recalibration intervals of measuring equipment.
ISO/IEC 17025:2017 for laboratory requirements, and ISO/IEC 17011:2017 for the definition of accreditation.
The NIST policy on metrological traceability.
The AIAG Measurement Systems Analysis (MSA) reference manual, for the gage study terms.

Tip: The definitions here are paraphrased. When a contract, a procedure or an audit finding turns on the exact wording, quote the source document itself.

Common questions

What is TUR in calibration?: TUR, the test uncertainty ratio, compares the tolerance of the item being calibrated with the expanded uncertainty of the calibration at about 95 percent coverage. For a tolerance of ±T it is T / U, as ILAC-G8 writes it. A tolerance of ±0.010 mm checked with an expanded uncertainty of 0.002 mm gives a TUR of 5:1. A low TUR calls for a guard band or a better calibration process.
What is the difference between calibration and verification?: Calibration measures how far an instrument reads from a reference and states the uncertainty of that result. Verification takes the result and decides whether the instrument meets a stated requirement, such as your tolerance. The VIM defines them as separate operations and warns against mixing them up. A calibration certificate can report results with no pass or fail at all.
What do as-found and as-left mean?: As-found is the reading an instrument gives when it arrives for calibration, before anyone adjusts it. As-left is the reading after the work is done. As-found tells you whether the instrument was measuring correctly during the interval that just ended. As-left shows it is fit to go back into service. If nothing was adjusted, the two are the same.
What is a guard band in calibration?: A guard band is the gap between a tolerance limit and a tighter acceptance limit that the pass or fail decision is made against. It is sized from the measurement uncertainty, so an instrument passes only when its reading sits far enough inside the tolerance. JCGM 106 defines the term. ILAC-G8 notes that a guard band equal to the expanded uncertainty is common.
What is the difference between accuracy and precision?: In the VIM, accuracy is how close a measured value is to the true value, and precision is how closely repeated readings agree with each other. A gage can be precise and still inaccurate if it repeats the same offset. On data sheets and in the shop, accuracy usually means a stated limit such as ±0.02 mm, and precision is often used loosely for accuracy or for the number of digits on a display.
What does NIST traceable mean?: It means the results on a certificate can be related to standards maintained by NIST through a documented, unbroken chain of calibrations, each with a stated uncertainty. The claim is about the results, not the instrument. NIST does not certify or approve traceability claims made by others. The calibration provider has to support the claim and you have to assess it.

Put it into practice

Got a gage list? Send it over and we load it for you, usually in a couple of business days. Free on every plan. Then log every calibration to a tamper-evident audit trail and produce records on demand.

Create your account

Axiospec is a documentation and workflow tool. It helps you keep clean, traceable, audit-ready records. Certification depends on your own processes, scope and assessor.