Correlation is not causation.
The classic example: that a bunch of storks appear around the same time a bunch of babies are born does not establish a causal relationship between the two events.
Image the reaction of a caveman faced with a functioning portable television. He might whack the screen and say "aha, the image was made by this area of the box." And he might whack the speakers and attribute the loss of sound to that part of the box. He might think when he killed the box, he killed the sound and the image. To him, the image and sound were "in" the box.
But that caveman knows nothing of information, signal processing, broadcasting. He is quite wrong. Correlation is not causation.
That's the same way I see science constrained by "methodological naturalism" trying to address "mind" "soul" and "spirit." Ditto for "world" and "reality." The domain of science is intentionally and significantly reduced, it does not have the means to discern beyond naturalism.
Post Hoc Ergo Propter Hoc
This is the most productive line of epistemology. How well do you know it?