Using observation for an unobservable issue

When we are dealing with data, we can do experiments to test whether what we observe matches the hypothesis. But we are dealing with the unobservable data of consciousness.

The best scientists can do is to correlate unobservable experiences with observable processes. For example, the feeling of hunger is associated with visible activity in the brain's hypothalamus.

But collecting correlations does not explain why conscious experiences correspond with brain activity.