If I had one word to define science I would quote experimentation.
Put the experiment you plan to do on the table and we will make science.
Repeatability
Observation is repeatable but does not mean that this is a fact from which you can deduct causality, generality, or anything else.
An experiment is a human design and action that produce an interaction of different elements of matter (from mineral to complex biological form of life). This experiment leads to results, which means data (from natural senses to sophisticated measurements) and these results must be evaluated at least about:
- their reality (objective measurements)
- their uncertainty (nothing is certain, but we can measure uncertainty.
An evidence-based result is repeatable, it is the aim of experiences.
Falsifiability
Experiences are repeatable if you put the same parameters in the experiment. And if an experiment is repeatable you can conclude that in this setting interaction of the matter leads to this result. If the design and the analysis of uncertainty are correct you can deduct some elements of a theory. But this does not exclude falsifiability of this theory when new facts resulting from other experiments will come and partially or mainly contradict this theory. Facts are rarely falsifiable, but theories are always falsifiable.
https://www.aier.org/article/fact-vs-faith-science-in-an-age-of-unreason/?fbclid=IwAR3qmS_C0scJRRvOVmHr7CSe7xoQiF-zCgFdFEodgTeaYJP-9NxQw6VUio8
https://www.amazon.com/Evidence-Things-Not-Seen-Reflections/dp/1942503628?fbclid=IwAR1uibqeE1OeBCVphOtfdnX4gLpR-oxlaKDqmJcvuuZ3EDrqqNXp95OUxPQ
