|Check your blood pressure. Is it high? Do you want to lower it? Do you then take pills under a physicians direction to do so? Now, those are the easy questions. Here is the hard one. Do the medications ("pills") that your physician has prescribed make a difference?
You would think this would be a no-brainer. Measure your blood pressure before you start taking the pills, and then afterwards. If you blood pressure is lower afterwards, then the prescription is working. If not, try something else.
Unfortunately, due to your daily variability in blood pressure, the decision making is not that simple. First, there must be enough measurements taken to derive a reliable average, and to gauge just how variable the data are. Then, the "two sigma" standard deviation lines need to be added to see if these limits about the average contain the right number of data points. Then, a "t-test" has to be done on the averages to see if they are in fact "statistically distinguishable" at the 95% confidence level.
That is what this tutorial lab is all about. The graph above shows these methods applied to one set of real data, using Excel as the analytical vehicle. In doing this, we learned that we could not reliably tell the efficacy of the first med used, but could reliably say the second combinations of meds was effective in lowering the person's blood pressure.