Same word, different worlds

Bias

Bias turns up in 3 fields: statistics, metrology, electronics. Here is what it means to the people in each one.

Engineering Statistics

Bias in statistics

Systematic offset between the average of repeated measurements and the true or reference value.

Bias is the difference between the mean of many measurements of a reference part and that part's accepted value. Unlike random noise it does not shrink with averaging, so it has to be found by measuring a calibrated standard and then removed by calibration or an offset. Linearity is bias that changes across the measuring range, such as a gauge reading true at the low end and 2 microns high at the top.

"There's a 1.2 nm bias between the two SEMs. Same wafers, same sites, tool B always reads high."

Measurement Uncertainty

Bias in metrology

The estimated systematic offset of a measuring system: its average reading minus the reference value.

An estimate of systematic error, obtained by averaging repeated measurements of a reference standard and subtracting the standard's assigned value. A torque analyzer averaging 50.6 N·m on a 50.0 N·m reference has a bias of +0.6 N·m. Labs either correct for bias or, when it is small and left uncorrected, carry it explicitly in the uncertainty budget so it cannot hide.

"Check standard shows a bias of plus 3 ppm since the move, apply the correction until we recalibrate."

Electronics & Semiconductors

Bias in electronics

A steady DC voltage or current applied to an electronic component to establish its resting operating point.

Bias sets the steady-state operating conditions of an active device like a transistor or vacuum tube before any AC signal is applied. By applying a specific DC voltage or current, the designer forces the component into a linear operating region, ensuring that incoming signals are amplified symmetrically without clipping. Incorrect bias leads to distortion, excessive power dissipation, or thermal runaway.

"The audio output sounds terribly distorted because the quiescent bias on the output stage is drifting too low."

More words with double lives