Quality & InspectionFree worksheetPDF, Excel, Word exportDiscrete Manufacturing

Gage Bias and Linearity Study Worksheet

A bias and linearity study checks whether a gage reads true across its working range. This worksheet computes bias at each reference point and shows it as a share of tolerance, so you can see at a glance whether error grows with size.

Download:

Reference part	Reference value (mm)	Average of readings (mm)	Feature tolerance (total) (mm)	Bias (avg - ref)	Bias % of tolerance
				0.003	1.5%
				0.002	1%
				0.005	2.5%
				0.008	4%
				0.011	5.5%

Mean bias

0.0058

Largest positive bias

0.011

Largest negative bias

0.002

Bias spread across range

0.009

A large spread suggests linearity error.

For each reference part, measure it at least 10 times, average the readings and enter the average. Reference values should come from a traceable calibrated source. Fit a regression line of bias against reference value in Excel for the full AIAG-style linearity assessment.

Automate bias and linearity checks from your gage data

Optional review of how to feed gage readings from your ERP or QMS into repeatable MSA analysis.

Optional. The tool and every export work without this. No spam.

How to run the study

Select five or more reference parts or masters that span the operating range of the gage. Have the appraiser measure each one repeatedly, ideally ten or more readings per part in random order. Enter the reference value, the average of the readings and the feature tolerance. The worksheet subtracts the reference from the average to get bias and divides by tolerance to express it in percent.

Use traceable reference values, not production parts measured on the same gage
Randomize the order of parts to avoid drift effects
Keep temperature and operator constant during the study

Reading bias and linearity

Bias is the difference between the observed average and the accepted reference value. Linearity is how that bias changes across the range. If bias is roughly constant, the gage has an offset that can be corrected. If bias grows with size, the gage has a scale or geometry problem. The spread row shows the difference between the largest and smallest bias, a quick screen before you fit a regression line.

Common mistakes

The most frequent error is using too narrow a range, which hides linearity problems. Another is using parts whose reference value was set by the same gage under test. Finally, teams often judge bias against the gage resolution rather than the tolerance. Judge it against the feature tolerance and your own acceptance criteria, and document the decision in the calibration or MSA record.

Frequently Asked Questions

How many reference points do I need?

Five points spread across the operating range is the common minimum for a linearity study. More points give a better regression fit. Include points near both ends of the range and at least one near the middle where most production parts fall.

What is an acceptable bias?

There is no universal limit. Many organizations compare bias to the feature tolerance and require it to be a small fraction, often under about 10 percent, or test whether it is statistically different from zero. Follow your customer requirements and internal MSA procedure.

Is bias the same as a calibration error?

They are related. Calibration compares the gage to a standard at a few points and may adjust it. A bias study evaluates the whole measurement process, including the appraiser and the part, against a reference value in actual use.

Can I export the results?

Yes. Use the export buttons to download the table as Excel or CSV, or as a PDF for inclusion in your MSA file.

If your gage readings already flow into SAP QM, SyteLine or another ERP, Netray can help pull them into this kind of analysis automatically.

Talk to an Expert Explore More Free Tools

Related Tools

Quality & Inspection

More Quality & Inspection tools

See all

Back to All Tools