It is pretty much how Quintessence explained. A matter of having method and proper background.
That is not really how things happen in science. Science knows how to deal cautiously with uncertainty, how to deal with confidence intervals, how to revise and improve theories and knowledge in general.
It is the very watered-down and often wildly misrepresented popular understanding that is called, with various degrees of accuracy and honesty, as being "science", that is often irresponsibly unreliable.
One issue I have with the way 'science' tends to be discussed is that it is treated as a normative concept, rather than the positive reality that results from it being a human endeavor.
All too often, at least in certain fields, 'science' (well scientists...) does not know how to deal with uncertainty, randomness and confidence levels. That is why there are so many false positives in many fields and woeful replication rates.
The quest for novelty, the need for scientists to be published, the hunt for funding, the need to protect ones reputation, poor methodology, poor understanding of statistics, poor methodology, wishful thinking, groupthink, confirmation bias and other factors all contribute to making the positive reality very different from the normative expectations in many of the sciences.
I think for society to maximise the benefits of the sciences and minimise the harms, it should be recognised that science is a human activity and that many of its failings are result from things that fall within the boundaries of what is considered 'proper science', not from dishonest manipulations of the concept.