Established science these days (since late 19th century) involves testable hypothesis, solid mechanical explanation and rigorous experiment when possible, reproducible too.

"Unestablished" "science" tends to skip one or more of these. A lot of "soft" sciences also skip a lot of steps before publishing findings.

The most commonly skipped one is describing experiments in sufficient detail for them to be possible to reproduce. The other is making unwarranted conclusions from a special group or situation.

Ideally your culture has evolved a common set of norms that enable it to be high trust. Another word for this is “religion”. Additionally, having a more similar than not genetic makeup may be necessary (the diverse society experiment is still ongoing).