Statistics deals with uncertainty. Figuratively, statistics breathes the air of uncertainty.
How ironic that scientists state definite conclusions, and science reporters report the definite conclusions, after they analyze their murky data statistically.
The proper way to state scientific conclusions after statistical analysis is to list the range of possibilities of their interpretation with their probabilities, along with obvious acknowledgements of possible deficiencies in the models.
So for example, the null hypothesis is rejected under alpha less than or equal to k, but is accepted when alpha is j is higher. This makes it clear that rejection or acceptance of an hypothesis is anything but crystal clear. Rejection or acceptance of any hypothesis is contingent on a decision that cannot be perfectly free of arbitrariness.
In this way, the result does not appear to the novice to be certain, when it is merely probable, and probable only if no errors have been made (including removal of outliers with insufficient justification), including no naïveté in the model construction.
What a laugh! Another definition of man: the animal who is smart enough to build sound scientific models, but never seems to muster the energy, the intellectual honesty and the courage required to do so.
Of course this is not done – man’s intellectual dishonesty won’t let it triumph.