sample
You cannot taste an entire pot of soup to check the seasoning — you take one spoonful and trust it speaks for the whole pot. That spoonful is a sample: a small, manageable portion taken from a much larger body of material so it can actually be examined. Analysis almost always works on a sample, not on the whole truckload, lake, or patient.
In analytical chemistry, a sample is the specific portion of material that is taken, prepared, and measured in order to learn about the larger 'population' it came from. The sample contains the analyte (the thing you want) sitting in the matrix (everything else). The reliability of your final answer can never be better than how well the sample represents its source.
It matters because the most careful instrument in the world cannot fix a bad sample: if your spoonful came only from the unsalted top of the soup, your verdict on the whole pot will be wrong. So a huge part of good analysis is sampling — choosing, taking, and storing samples so they truly stand in for the material of interest.
To check a river for pollution, an analyst does not test the whole river. They collect a few liters of water at chosen spots and times — those collected portions are the samples that stand in for the river.
A sample is a small portion meant to represent a much larger whole.
Keep three ideas distinct: the sample is the material you took, the analyte is what you measure in it, and the matrix is everything else in the sample. They are easy to blur but mean different things.