A good scientist's conclusions are never certain. They, instead, deal with theories, postulations, principles, and many other assumptions that are not proven, but certain enough to be taken as fact. Unless these assumptions are made, science can't advance.
If we agree to a set of foundational assumptions, we can derive "truth statements" using those assumptions -- that's the basis of the "truth" of epistemology and mathematics. All undeniable truth statements in these approaches are predicated by the implicit, "if we agree that these foundational assumptions are true..." condition.
You may believe that you actually exist, but you really may not. You may just be a dream (solipsism).
Logic gives us a cleaner system that allows for undeniable conclusions, such as a tautology:
Example: "If neither John nor Sue is here, then John is not here."
And rule of replacement:
p º ( p Ú p ), p º ( p · p )
Example: "Paul is tall," is equivalent to, "Paul is tall and Paul is tall."
On paper, you can always argue that if A=A then A, with respect to itself, is A. Real-world science doesn't allow for this sort of clean analysis, and bold conclusions can cost a scientist their academic reputation.