Measurement Systems Analysis

Probable Error Analysis

Estimate measurement-system probable error from repeated measurements and evaluate whether the selected measurement increment is appropriate. Probable Error provides a practical view of measurement uncertainty and discrimination capability, complementing traditional Gage R&R and EMP-style studies.

Relation to Gage R&R: Gage R&R commonly decomposes measurement variation into repeatability and reproducibility components. Probable Error focuses on the expected size of measurement uncertainty and whether the measurement resolution is fine enough to support practical decisions.

Required Data Structure

Use long-format data with repeated measurements for each part or sample.

Part column: part or sample identifier
Measurement column: numeric measured value
Each part must have the same number of repeated measurements
Subgroup size must be between 2 and 25

Download Example CSV

Part	Measurement
P1	10.005
P1	10.035
P2	10.333
P2	10.368
P3	10.666

Upload and Analyze

CSV file

Part column

Measurement column

Measurement increment

Upload a CSV or load the example study to begin.

Probable Error Results

Interpretation

The acceptable increment band shown here follows the implemented rule: the measurement increment should fall between 0.2 × PE and 2 × PE. Values below the lower bound suggest unnecessary display precision; values above the upper bound suggest inadequate measurement resolution.