When you see someone uses numbers and percentages while arguing about something, you need to ask yourself, and ask them if time and place allow, first to give you the exact address of the survey or research or other forms of publications where these numbers are coming from. Second, we all need to know that some percentage says very little, even despite being large, sometimes of any association between two factors, issues, or phenomenon.
Statisticians apply complex mathematical and econometric models over “large” and multi-faceted data just to be content with some ceteris paribus relationship between two factors, meaning everything else constant (i.e. ignored). Then they use other complex mathematical tests to reject themselves, within some statistical confidence, e.g. 95% (by the way, that doesn’t mean they would be 95% sure). Only then, they may say there might be some correlation, and not causation, between the two factors.
Let me put it this way, it’s way more efficient to say your opinion about something if you have the courage and not be afraid of being incorrect. Do not be tempted to deceive people with numbers. That’s fraud.
No comments:
Post a Comment