Human sciences like psychology often utilize experimentation. However, certain disciplines such as economics and political science find it challenging to conduct controlled experiments.

Fun Fact: Why is this? Well, it's mostly because these disciplines deal with complex systems, where controlling every variable is impossible. Think of trying to control every single factor in a country's economy - pretty impossible, right?

But there's a twist - the knowledge we gain from these experiments can help predict policy outcomes and reforms.

Experiments do hold value, as shown by psychology and behavioural economics through randomized controlled trials. But, it's like walking on thin ice - experimental evidence can sometimes be misunderstood, and it might even give us a false sense of certainty.

Example Time: Remember the disastrous effects of over-relying on mathematical models in economics? Yeah, that's what we're talking about.

Moreover, sampling issues like the WEIRD bias (Western, Educated, Industrialized, Rich, and Democratic societies' overrepresentation) can muddy the water, causing problems in our knowledge pool.

Here comes a story of rats, drugs, and parks! Bruce Alexander’s “rat park” study revolutionized how we perceive addiction - shifting blame from individual's moral and mental failings to social dislocation. This knowledge, however, was ignored while governments persisted with their "war on drugs."

Real-World Example: It's like if you're given a banana but keep asking for an apple, even when the apple's rotten.

Despite attempts to replicate the rat park study's results and some concerns about the experiment's design, there's now a broader (though not universal) acceptance that we need more compassion for people struggling with addiction.