Scientific methodologies that people just don't get

I think most of these probably apply to many other sciences, especially :

Psychological research is largely based on statistical findings. When a researcher obtains a finding in a study that shows as “significant,” he or she could be wrong — in fact, the finding could not be a valid representation of how things are in the world, in spite of what the researcher found. If a researcher finds something to be significant, but is actually wrong, we call this “Type-1 Error.” When a researcher obtains a finding in a study that shows as “not significant,” he or she could be wrong. If a researcher finds something to be not significant, but is actually wrong, we call this “Type-2 Error.”

Though we don't use those terms in astronomy.

A true experiment is the only way that we can make causal inferences regarding the relationship between variables. And a true experiment requires random assignment to different conditions. If you are doing a study and want to see if people in one situation (e.g., those who drink a lot of coffee) behave differently from those in another condition (e.g., those who do not drink coffee), you need to randomly assign people to one of these two conditions. A study that looks for differences in some outcome between naturally occurring groups (e.g., people who regularly drink coffee versus those who don’t) is a “quasi experiment” and cannot establish whether changes in one variable cause changes in the other variable.

That's the classic explanation as to how you establish causation if correlation isn't enough : you change one and only one variable and see what happens. Outside of a lab, you'll never get these precise conditions, so you can only make reasonable inferences.

Sometimes, a variable will have one effect for one group of people and a totally different effect on some other group of people. Thus, the effects of some variable on some outcome variable often depend on some other variable. If you want to see whether teenagers are more likely than other age groups to choose to be in the company of others, it would be helpful to know whether they have, for example, combed their hair within the last 24 hours, and whether they will be in the company of family members or strangers. Teens who have not combed their hair might be more likely to choose the company of others if they are going to be in the company of family members.

People often fall into the trap of thinking that there is a single cause of any behavioural outcome. In fact, trained psychologists know full well that multiple causes are often at play in shaping any behaviour. Mass shootings, which comprise a horrific facet of modern life, provide a clear example of this kind of reasoning. Some people argue that mass shootings are the result of mental health problems. Others argue that mass shootings are the result of poor laws regarding gun control. In fact, both issues have been shown to play a role. And several other factors have also been implicated in this complex societal issue. For the lion’s share of behaviours, multiple factors are at play.

https://www.psychologytoday.com/us/blog/darwins-subterranean-world/201711/10-key-psychological-concepts-many-people-just-don-t-get