Until someone is shown that they are wrong they are highly unlikely to suspect they are wrong. We've used all of our evaluation tools to formulate our original hypothesis, it seems ridiculous to believe we could use the same tools to find out where our hypothesis is wrong. Using the same faulty tools will give us the same faulty response.
This is one of the key flaws in human cognition - we tend to establish a hypothesis and then perform tests that would
confirm the hypothesis,
not tests that would
falsify the hypothesis.
We can be trained out of that mental trap, but it can be difficult, and more so if the hypothesis in question is emotionally significant.