Building a good calibration curve is a critical point when applying a quantitative method. As you know, in order to know if the calibration is suitable we can check out (among others) the R2 value. But, be careful, we tend to overestimate the importance of this parameter. Look at the following figure called the Anscombe’s quartet. All the dataset have the same mean value for x and y and… the same R2! Thus, as you can see, graphs are very useful to study our calibration curves (the residual graphs can be very useful too).
Here you can see another (and funnier) picture that shows us how important graphical representation is (the statistics are always the same).
Sometimes, building calibration curves is complicated… Should we include the intercept? How can we calculate the Limit of Quantitation? In this course you should use the concepts you learnt in the “theoretical lessons”, but here you can find a nice series of articles by John Dolan that might help you: 1, 2, 3, 4 and 5.
In the future, you might learn that there are more fitting methods than the linear one (quadratic or cubic ones for instance), but in the meanwhile just take it with a bit of humor.