matrix effect
A matrix effect is when everything else in a sample — all the stuff that isn't the thing you're measuring — quietly changes the answer the instrument gives. Imagine trying to hear one person in a noisy café: the chatter around them doesn't change what they say, but it changes how well you catch it. The other ingredients in a sample can do the same to a measurement, nudging the reading up or down.
More precisely, the matrix is everything in the sample besides the analyte, and a matrix effect is any influence that background has on the measured signal. The same amount of analyte may read high in blood, low in seawater, and correct in pure water — not because the analyte changed, but because the surrounding salts, proteins, fats, or ions altered how the instrument responds.
Matrix effects are one of the most insidious sources of error because the analyte itself is perfectly fine — the bias hides in the background. The usual cures all involve making the standards and the sample share the same surroundings: matrix-matched calibration, the method of standard addition, or sample preparation that strips the matrix away before measurement.
A drug measured by mass spectrometry reads 30% low in blood plasma but correct in pure water — co-eluting plasma compounds suppress the signal, a textbook matrix effect cured by standard addition.
The background, not the analyte, skews the reading.
Don't confuse a matrix effect with an interferent. An interferent produces a signal of its own that adds to or masks the analyte's; a matrix effect changes how the instrument responds to the analyte itself — suppressing or enhancing its signal — without necessarily giving any signal of its own.