There is a lot of things to say about that question. Starting by the nature of the "aleatory" in the real world not so behaved as in Excel. The quality of the function used to create random numbers. The meaning of significance level and the effect of the number of samples in the statistical tests. In the reproducibility as part of the scientific method and so on. But in truth, the great conclusion you have is real: "inferences and conclusions in our researches could be meaningless", or not !

I believe that the real world is subjected to nature's laws that is why we expect to achieve suggestive, reasonable and SIGNIFICATIVE patterns from real data colected according to an appropriate and comprehensive sampling design.

Although the point here is that the same kind of patterns seems to arise from simulated and aleatory data.

This really make my doubt about some inferences and conclusions achieved in many serious scientific research.

Daniel. Excel aleatory data gives several values over a limited line. If you create a non linear function for that sample, then you may simulate the pattern of that particular function.

For example, an unbiassed  6 cubic die is made of faces with values  1,2,3,4,5,6 (linear). But what would happen if instead of that, faces are 0, 10, 10, 20, 90, 90 ?

The answer is that the two dice have different patterns, or distribution laws. So they produce different aleatory samples that must produce different interpretations.

It seems that nature and individuals behave most of time in non-linear patterns.

Emilio

Although your answer could sounds reasonable at a first glance, I believe that individuals behave in multilinear patterns, according to the uncountable variables present in our known universe