Originally Posted by ForestryProf
Need to chime in here. I freely admit that my experience with straights is horribly lacking compared to present company; however, being a research scientist, I do have some experience with experimental design. Joe is correct. Stropping before and after leaves the experiment completely confounded with regard to the causal agent of the effect (i.e. is it the extra stropping, or the time that the stropping was done that is causing the difference). On the other hand, if you don't care which of the two factors is most affecting the razor, than it does not matter that the experiment has confounding. This may create an additional problem, if the effects are equal and opposite, you will find no difference even though each of the two variables is affecting the razor. For instance, if the extra stropping is good, and stropping following a shave is bad, you may find no cumulative effect.
If you want to get at causality, you actually need to do much more than Joe suggests, because you also have a time element involved. With each variable added, you need to double the number of razors to address the issue. More trouble still--you are assuming that there is no experimental error. Without replication, you really cannot be assured that there is not variability among razors, shaving technique, stropping, initial hone, etc...
God I love my job ;-)
Ed