Exercises 6–8: the implication of statistically significant diff values
keep in mind that for reasons mentioned earlier, the randomization distribution above is demonstrating what is likely to happen by chance alone if the treatment was not effective.
as stated in the previous lesson, you can use this randomization distribution to assess whether or not the actual difference in means obtained from your experiment (the difference
between the mean weight of the 5 actual control group tomatoes and the mean weight of the 5 actual treatment group tomatoes) is consistent with usual chance behavior.
the logic is as follows:
 if the observed difference is "extreme" and not typical of chance behavior, it may be considered statistically significant and possibly not the result of chance behavior.
 if the difference is not the result of chance behavior, then maybe the difference did not just happen by chance alone.
 if the difference did not just happen by chance alone, maybe the difference you observed is caused by the treatment in question, which, in this case, is the nutrient. in the context of
our example, a statistically significant diff value provides evidence that the nutrient treatment did in fact yield heavier tomatoes on average.