What a learner can do afterwards
- Prepares a calibration series and states why the standards must bracket the sample
- Explains what the blank corrects for and what it cannot correct for
- Chooses between external calibration, internal standard and standard addition for a stated sample
- Names a matrix effect and explains how it biases a result even when the calibration line looks good
1 · Read
An instrument reports a signal, never a concentration. You turn one into the other with standards: known samples spanning the range, bracketing your unknown from both sides, joined by a calibration line you read the sample off. Treat strange points with discipline. One far off point can drag the slope and poison every reading, so check how it was made and drop it only on proof of a concrete mistake, stated openly.
The blank is the same run with everything except the thing you measure. It corrects for background signal from reagents and the instrument itself. Know its limits: it cannot correct for effects that change how the analyte responds, nor for a sample sitting outside the tested range.
Pick the calibration method your sample demands. Clean samples suit external calibration against separate standards. Loss prone preparation calls for an internal standard that rides along and corrects. A tricky matrix like seawater demands standard addition, where standards are added into the sample itself. For a metal in seawater, the blank is the same seawater matrix minus the metal, and standard addition beats the alternatives because it carries the matrix along.
Name the quiet killer: the matrix effect. The sample matrix shifts the instrument response, biasing results even when the calibration line looks beautiful. When bias survives good lines and good blanks, suspect the matrix and switch to standard addition.
Bracket with standards, subtract a matched blank, pick the method the matrix demands.
More to read
Outliers
When you fit a line through a set of measurements, most points sit close to it. Some points sit far away. A point like that is called an outlier. The distance from a point to the line, straight up or down, is called the residual, or error. An outlier has a very large residual.
Outliers matter because they can pull a fitted line toward themselves. If you are drawing a calibration line through your standards, one strange point can change the slope a little, and that change follows the sample you later read off the line.
So what do you do when you see an outlier? First, look at it closely. Ask how that number was made. Sometimes the point is simply wrong data. A digit got typed wrong, a standard got swapped, or something spilled into the cuvette. If you can show the point came from a mistake, you leave it out and say why.
Other times the point is real. The instrument really did respond that way. Real outliers can reveal something important, like a standard outside the range where the method behaves, or a sample that does not behave like the others. Then you keep the point and investigate.
The rule of thumb: never delete an outlier just because it is inconvenient. Check it, explain it, and only then decide. A trustworthy line is one where every point on it can be accounted for.
2 · Watch
Take it off screen
Where it sits
Optional connections
8 questions wait behind this lesson, each with its answer explained. Every answer feeds the sky: stars light as they are learned, and dim when it is time to come back.