It would take 7 packages each with 37 days extra (the largest extra delay observed) to explain the 3 days average extra shipping time..
First, I'm not saying that the vast majority of the packages where delivered on time as it is really hard to get a three day difference like that without a significant quantity of packages being delivered after the plain taped packages. Also, considering that the authors didn't mention that there were a number of packages arriving outside of that three day delivery window, I'm assuming that most of them arrived within that three day delivery window. Thus, the following would work to give you an average of 3.025 days:
1 Day - 5
2 Days - 42
3 Days - 12
4 Days - 20
37 Days - 1
However, that also gives a standard deviation of 3.933 days which implies that the outlier is having a significant impact upon the results. Removing that outlier (dropping it completely) gives us an average of 2.59 days and a standard deviation of 0.93 which is in line with what we would expect. Granted this would still clearly indicate that there is a discrepancy from the plain tape package.
You do know that outliers are samples that are far from the population, right? If there are 7 or more out of 80 or 89, then they are not far from the population. At that point they are representative of the population.
Yes, I know that and I believe that I just demonstrated why it is important to know the standard deviation when discussing this data as well.
At what point do you admit that you arent actually educated enough to have been equipped with the skills necessary to make reasonable arguments?
Likely around the same time that you admit that I might actually know what I'm talking about, although I if I mention the fact that I work with hidden Markov models will you at least give me the benefit of the doubt that I've had a couple math courses in my day?