Explanation:
The scientific process is supposed to be something like ...
- make observations
- formulate a theory that explains the observations
- use the theory to formulate experiments and predict expected additional observations
- seek to determine if the predictions are correct, especially seeking out data that might show otherwise
- modify the theory as needed to explain all observations and data
At any point, the observations and theories can be published for critique by other observers, experimenters, and scientists.
The theory should be sufficiently defined so that it is possible to differentiate a relevant experiment or observation from one that is irrelevant. Similarly, every effort should be made to separate causality from correlation, and to identify or rule out alternate causes or additional causes from the one(s) theorized. Experiments should be sufficiently well-defined so that they can be reproduced.
__
Additional comment
These days, a common error is to ignore or discount observations that don't agree with accepted theory, including observations by "lay persons." Such "scientists" fail to remember that science is only advanced by attempting to explain the unexplained.
