Somebody said once that all statistical data can be manipulated to prove someone else's point of view. And this statement was true because data can be collected in such way to accept either Null Hypothesis or Alternative Hypothesis (or either one –there are four cases available) therefore it is not about data interpretation as it is about formulating the most appropriate Null Hypothesis.
Here are some facts:
1. when politicians select the wrong Null Hypothesis, we all get into trouble as wars, famine, unemployment!
2. when banks select the wrong Null hypothesis, we all get into trouble as “lost money” across the globe
3. When pharmaceutical companies select the wrong Null Hypothesis, people get the wrong pills to the wrong body
4. When we decide that people are superior to animals we choose the wrong Null Hypothesis
What do you think that we miss to set-up the correct Null Hypothesis that would imply that everybody is happy? Because if we set-up the Null Hypothesis as everybody should be happy, we can then work on collecting data to ensure sure that all systems are in place to achieve it!