Old me would have disagreed with this, but it seems true. Scientists create and test models to better understand the state of things. They must then convince others in their field of the value of those models. The field must then convince the public. https://www.scientificamerican.com/article/the-idea-that-a-scientific-theory-can-be-falsified-is-a-myth/

There are models that are easy to convince. In computer science, the speed of "QuickSort" vs "Bubble Sort" isn't in debate. We create models on top of models to give us quicker ways to convince ourselves. p-values are some such example in some fields.

So many questions in science and research don't have objectively correct answers. They have answers that are accepted or seen as reasonable. Is a study of 100 or 1000 participants reasonable to draw a conclusion? What's a reasonable level of "representative"-ness?

What we're left with is the consensus of experts. An expert is someone who's recognized as an expert. Expertise is thus socially constructed, and the scientific claims of a field are thus also socially constructed. What the public accepts as credible is then what matters.

Of course, there's logic and reason to all this judgement of credibility and expertise. If a field repeatedly puts out things considered wacky and not matching people's perception of reality, it will lose credibility. The same for judging expertise of researchers in a field.

This is kind of tough for me to write knowing where I'm coming from on science. I've strenuously defended "the scientific method". The fact is, the scientific method is whatever works in a field to gain research credibility within the field and in the outside world.