To be confident that a cause-and-effect relationship exists, it is necessary to

To confirm a cause-and-effect relationship one must perform a controlled experiment.

Naturalistic observation only describes behaviour, it cannot isolate causes.

Developing a positive correlation shows association but never proves causation.

Testing for a negative correlation likewise reveals relationship, not cause.

Only manipulating an independent variable while controlling others yields causal proof.

Random assignment in the experiment rules out confounding explanations.