During my employment in Science lab a chemist, I was a member of a research that required collecting and evaluating a considerable total of information pertaining to the human mind as we were attempting to create a vaccine. We were so mixed up at that time because all my further experiments depend upon that result and then we began to guess about it. We reprised the whole experiment again just to attain certain that there is no human error in our stamping ground and we had to manage the analysis again which may forbid it.
And so it got very clear that a dissimilar case of analysis would have been more pertinent to the problem and would have given results that supported our theory. We changed the regression technique within the model and re-ran the data.
I would have concentrated mainly on the robustness of the information we accumulated and then progress to a correct analysis on which our whole research depended.