It's been trending down for a while. Tons of "soft science" fields have problems. The psychology field, for instance, has a replication crisis, where a lot of major "discoveries" in recent decades either aren't replicating at all, or shown to have negligible real-world effects in followup studies. Stereotype threat is a major one, as is Implicit Bias. Not coincidentally, both of these supposed effects are used by Social Justice types to justify their activism, and also coincidentally, researchers that point out the problems with these concepts are often smeared by activists and others in the field.

P-hacking is another problem that exists in a lot of fields. The fields with integrity are trying to fix this, with methods such as pre-registering studies, but there are fields like psychology that have some pretty fundamental problems.