internal standard
An internal standard is a known amount of a chosen reference substance you deliberately add to every sample and every standard alike, then track alongside your analyte. Think of two runners starting together: if both are slowed by the same wind, you can still judge their relative speed by how far apart they finish, even though neither ran their true pace.
In practice you mix in a fixed amount of a compound that is not your analyte but behaves similarly in the instrument. Instead of reading the analyte's signal on its own, you read the ratio of the analyte's signal to the internal standard's signal. Because both rise and fall together when something goes wrong — a little lost in transfer, a slightly different injection volume — their ratio stays steady.
It matters most where small, unavoidable losses or fluctuations plague a method, as in chromatography and mass spectrometry. The internal standard cancels out errors that would otherwise wreck precision. The catch: it must be absent from the original sample, well separated from the analyte, and similar enough to share the same ups and downs.
In a gas chromatography assay for ethanol in blood, n-propanol is added to every sample at a fixed amount. The lab reports the ethanol-to-propanol peak-area ratio, so a slightly small injection no longer biases the result.
Reading a ratio, not a raw signal, cancels shared losses and fluctuations.
Do not confuse internal standard with standard addition. An internal standard is a different substance added to track losses; standard addition adds more of the analyte itself to correct for matrix effects. They solve different problems and are sometimes combined.