A useful question is:

Can we trust the measurement before we trust the analysis?

MSA helps answer that question. Because MSA is essential it is required in several quality systems. Automotive requires it in TS16949 in the measurement systems analysis manual. VDA requires it in VDA 5 Capability of measurement Process.  Later this year these 2 manuals will be integrated and for example  Aerospace (SAE) requires it in the RM13003.

What is Gage R&R?

Gage R&R, also written as Gauge R&R, GR&R, or Gage Repeatability and Reproducibility, is a statistical study used to evaluate the variation associated with a measurement system.

Gage R&R focuses primarily on two components:

Repeatability

Repeatability measures variation when the same operator measures the same part repeatedly using the same measurement equipment.

It is sometimes described as equipment variation.

For example, an operator measures the diameter of the same component several times using the same micrometer.

If the measurements are:

25.01 → 25.02 → 25.01 → 25.03 → 25.02 mm

the measurement system is showing relatively little variation.

If the results are:

24.92 → 25.07 → 24.98 → 25.11 → 24.95 mm

there may be a repeatability problem that needs investigation.

Possible causes include:

• Gauge condition
• Gauge design
• Gauge resolution
• Fixture movement
• Measurement technique
• Part positioning
• Excessive measurement force
• Temperature
• Vibration
• Surface condition
• Method or equipment limitations

What is Gage Reproducibility?

Reproducibility measures variation between different operators, appraisers, or measurement conditions.

For example, three operators may independently measure the same group of components.

If Operator A consistently obtains measurements around 25.02 mm, Operator B obtains 25.10 mm and Operator C obtains 24.95 mm, the measurement system may have an operator-related reproducibility problem.

Reproducibility is often associated with appraiser variation.

Potential causes include:

• Different measurement techniques
• Inadequate operator training
• Different interpretation of the measurement procedure
• Difficult-to-use equipment
• Poor fixture design
• Inconsistent part positioning
• Differences in measurement force
• Differences in how operators read or record results
• An inspection method that is not sufficiently standardized

Importantly, a reproducibility problem is not necessarily caused by the operators themselves.

Sometimes the measurement system is simply too dependent on operator technique.

A good MSA study therefore helps identify the underlying cause rather than simply blaming the appraiser.

Repeatability vs. Reproducibility

Does a measuring system report the same results when two or three operators measure the same part/ characteristic “blindly” several times? Between-operator reproducibility or between gage reproducibility is the other dimension of variation in a measuring process. Gage reproducibility is also referred to as appraiser variation, because it most significantly reflects on operators’ consistency in using the measurement system. Solutions to problems of reproducibility often involve operator training.

The easiest way to understand Gage R&R is to separate the two questions.

Question	Measurement variation
Does the same operator get the same result repeatedly?	Repeatability
Do different operators get the same result?	Reproducibility

Together, these components form the basis of Gage R&R.

A measurement system can have:

• Good repeatability but poor reproducibility
• Poor repeatability but good reproducibility
• Problems with both
• Relatively low measurement variation overall

Understanding which component is responsible is often more valuable than simply knowing whether the overall Gage R&R result passes or fails.

How does a Gage R&R study work?

A typical variable-data Gage R&R study uses a structured experimental design involving:

• Multiple parts
• Multiple operators
• Multiple measurement trials

A commonly used study design might involve 10 parts, 3 operators and 2 or 3 repeated measurements, although the appropriate design depends on the application and the measurement process.

The parts should represent the range of variation that the measurement system is expected to encounter.

Operators should also be representative of the people who normally perform the measurement.

Measurements are typically collected in a controlled and, where appropriate, randomized manner so that operators are not simply remembering their previous results.

The resulting data can then be analysed to determine how much variation is attributable to:

• Part-to-part variation
• Repeatability
• Reproducibility
• Operator effects
• Operator × part interaction

Overall measurement-system variation.

What does a Gage R&R actually tell you?

A Gage R&R study answers a fundamental question:

How much of the variation in our measurements comes from the measurement system rather than the parts being measured?

This distinction is critical.

Suppose your measurement data show significant variation.

There are two very different possibilities:

Scenario 1 — The process is genuinely variable

The measurement system is capable of distinguishing the parts accurately, and the observed variation primarily represents real differences between parts.

This gives you confidence that your SPC or capability analysis is telling you something meaningful about the manufacturing process.

Scenario 2 — The measurement system is variable

The measurement system itself contributes a large proportion of the observed variation.

In this situation, improving the manufacturing process may not solve the underlying problem.

You may first need to improve the measurement system.

Understanding %GR&R

One of the most commonly discussed results from a Gage R&R study is %GR&R.

This expresses the amount of measurement-system variation relative to the relevant total variation or study variation, depending on the analysis and reporting convention being used.

Commonly used guidelines (TS16949) are:

Less than 10%

The measurement system is generally considered acceptable for many applications.

10% to 30%

The measurement system may be acceptable depending on the application, measurement criticality, cost and consequences of measurement error.

Greater than 30%

The measurement system is generally considered unacceptable and should normally be investigated and improved.

However, these values should not be treated as an automatic pass/fail rule for every situation.

The consequences of measurement error matter.

A measurement system used for a critical safety characteristic may require considerably more scrutiny than a measurement system used for a low-risk characteristic.

The correct interpretation should therefore consider:

• Measurement purpose
• Product tolerance
• Process variation
• Risk
• Customer requirements
• Measurement resolution
• Industry requirements
• Cost of measurement error

---

What is Number of Distinct Categories (ndc)?

Another important Gage R&R result is Number of Distinct Categories, commonly abbreviated as ndc.

ndc provides an indication of how effectively the measurement system can distinguish different levels of part-to-part variation.

In simple terms:

Can the measurement system actually tell the difference between parts?

A measurement system may produce repeatable measurements but still lack sufficient discrimination to distinguish meaningful differences between parts.

An ndc of 5 or greater is commonly used as a practical guideline for an acceptable ability to distinguish categories of parts.

This is why looking only at %GR&R can be misleading.

A good MSA interpretation should consider the complete set of relevant results rather than focusing on one number.

---

ANOVA Gage R&R

There are several approaches to performing a Gage R&R analysis.

One widely used approach is ANOVA (Analysis of Variance).

ANOVA-based Gage R&R can provide a more detailed understanding of the sources of variation in the measurement system.

Depending on the study design, it can help evaluate effects such as:

• Part variation
• Operator variation
• Repeatability
• Reproducibility
• Operator × part interaction

The operator × part interaction can be particularly useful.

For example, an operator may measure some parts consistently but have difficulty measuring others. This can indicate that the measurement method, fixture, part geometry or inspection procedure is creating an interaction between the operator and the part.

This type of information can help you move beyond simply asking:

“Did the Gage R&R pass?”

and instead ask:

“Why is the measurement system producing this amount of variation, and what should we change?”

---

Average & Range Gage R&R

Another commonly used approach is the Average & Range method.

This method uses ranges and averages from repeated measurements to estimate the contributions of repeatability and reproducibility.

It can be useful because it is relatively straightforward to understand and implement.

However, an ANOVA approach can provide additional information about the structure of the variation.

The appropriate method depends on the measurement system, study objectives, applicable requirements and the level of analysis required.

---

Gage R&R is only one part of MSA

A common misconception is that MSA and Gage R&R are the same thing.

They are not.

Gage R&R is an important component of MSA, but a complete Measurement Systems Analysis can include several different studies.

For variable measurement systems, important characteristics can include:

Repeatability

Can the same operator obtain consistent results using the same equipment?

Reproducibility

Can different operators obtain consistent results?

Bias

Is the measurement system systematically offset from an accepted reference or true value?

Linearity

Does measurement-system bias remain consistent across the operating range, or does accuracy change depending on the size of the measurement?

Stability

Does the measurement system remain consistent over time?

Resolution / discrimination

Can the measurement system distinguish sufficiently small differences between parts?

Different applications may require different combinations of these studies.

---

Accuracy vs. Precision in MSA

MSA discussions often distinguish between accuracy and precision.

Precision

Precision concerns the consistency of measurements.

It is associated with characteristics such as:

• Repeatability
• Reproducibility

Accuracy

Accuracy concerns how close measurements are to an accepted reference value.

It is associated with characteristics such as:

• Bias
• Linearity
• Stability

A measurement system can therefore be precise without being accurate.

For example, a gauge might repeatedly measure a reference part as 10.20 mm when its accepted reference value is 10.00 mm.

The measurements are consistent—but consistently wrong.

This is why a Gage R&R study does not necessarily answer every question about measurement-system accuracy.

---

When should you perform a Gage R&R study?

A Gage R&R study can be appropriate whenever measurement data will be used to make important decisions about a process or product.

Typical situations include:

• Introducing new measurement equipment
• Introducing a new inspection method
• Starting a new manufacturing process
• Before a capability study
• Before relying on SPC data
• During PPAP activities
• When required by a customer
• When operators produce different inspection results
• When measurement results appear inconsistent
• When a process appears more variable than expected
• When changing measurement equipment
• When changing fixtures
• When changing inspection procedures
• When investigating measurement-related nonconformities

It can also be valuable when a measurement system has been in use for years.

Calibration alone does not prove that a measurement system is suitable for every application.

Calibration establishes the relationship of an instrument to a reference standard under defined conditions. MSA looks more broadly at how the complete measurement process performs in its actual application.

---

What happens when a Gage R&R fails?

A failed Gage R&R is not the end of the analysis.

It is a signal that the measurement system needs investigation.

The next step should be to determine what is driving the variation.

Depending on the results, improvements may include:

Improve the measurement equipment

The gauge may have insufficient resolution, excessive wear, poor design or inadequate performance for the application.

Improve the fixture

A poorly designed fixture can make consistent measurement difficult.

Standardize the measurement method

Operators should have a clearly defined procedure for positioning, clamping, measuring and recording results.

Train operators

Where reproducibility is related to differences in technique, targeted operator training may reduce measurement variation.

Improve resolution

If the gauge cannot adequately discriminate between relevant part differences, a higher-resolution measurement system may be required.

Control environmental conditions

Temperature, vibration, humidity or other environmental factors may affect measurement results.

Improve part presentation

Sometimes the problem is not the gauge but the way the part is located, oriented or held during measurement.

Repeat the MSA study

After improvements are made, a new study can determine whether the measurement system has improved sufficiently.

---

What if my Gage R&R result is poor?

If your Gage R&R is above the desired level, don’t immediately conclude that the gauge needs to be replaced.

The first step is to understand where the variation is coming from.

For example:

High repeatability variation may point toward the equipment, fixture, resolution, measurement environment or measurement method.

High reproducibility variation may point toward differences in operator technique, training, interpretation or the way the measurement is performed.

High operator × part interaction may indicate that certain parts or geometries are difficult to measure consistently.

This diagnostic approach is one of the most valuable aspects of MSA.

The goal isn’t simply to obtain a better Gage R&R percentage.

The goal is to create a measurement system that provides reliable information for making quality and process decisions.

---

Attribute MSA and Attribute Agreement Analysis

Not every measurement produces continuous numerical data.

Sometimes inspectors classify a product as:

• Pass / fail
• Accept / reject
• Good / bad
• Defective / non-defective

In these situations, a traditional variable-data Gage R&R may not be appropriate.

An Attribute MSA or Attribute Agreement Analysis can be used to evaluate the consistency of an attribute inspection system.

The analysis can investigate questions such as:

• Do inspectors agree with each other?
• Does an inspector agree with themselves over repeated evaluations?
• Do inspectors agree with a known reference?
• Are certain types of defects harder to identify?
• Is the inspection method sufficiently consistent?

This is particularly important when human judgement plays a major role in inspection.

---

Common Gage R&R mistakes

A Gage R&R study can produce misleading conclusions if it is poorly designed.

Some common mistakes include:

Using the wrong parts

The selected parts should provide an appropriate representation of the process variation relevant to the study.

Telling operators which parts they are measuring

This can introduce bias and reduce the value of the study.

Using an unrealistic measurement procedure

The study should reflect how the measurement system is actually used in production.

Using poorly trained operators

The study should normally involve representative users of the measurement system.

Ignoring the measurement environment

Temperature, vibration, lighting and other conditions can influence measurement results.

Looking only at %GR&R

A single percentage does not tell the entire story.

The complete analysis should be considered, including repeatability, reproducibility, part variation, interactions and discrimination where applicable.

Treating calibration as equivalent to MSA

A calibrated instrument can still be unsuitable for a particular measurement application.

Automatically replacing the gauge

A poor result should lead to investigation first. The underlying cause may be the method, fixture, environment, operator technique or measurement design.

---

MSA, SPC and Process Capability

MSA is closely connected to statistical process control and capability analysis.

The sequence is important.

You first need confidence that your measurement system can reliably detect the variation you are trying to control.

Only then can you confidently use that data for:

MSA → SPC → Process Capability → Process Improvement

For example, if a measurement system contributes substantial variation, a capability study may produce misleading results.

You could conclude that a process has a poor Cpk when the real issue is the measurement system.

Conversely, excessive measurement variation can sometimes mask meaningful process changes.

A capable measurement system therefore provides the foundation for trustworthy statistical analysis.

---

How we can help with Gage R&R and MSA

Performing a Gage R&R calculation is only one part of an effective Measurement Systems Analysis.

The real value comes from designing the study correctly, interpreting the results and determining what to do next.

Datalyzer support with MSA training and with our web based Gage management software.

We offer a free 2 months trial where you can perform MSA studies and see how MSA would work in your company. To request a trial go to the request for demo page