- Registration time
- 2009-12-10
- Last login
- 2018-2-4
- Online time
- 582 Hour
- Reading permission
- 70
- Credits
- 10231
- Post
- 2113
- Digest
- 0
- UID
- 328571
 
|
Yet, why so many rats, including yourself think you understand climate change better than Scientific America magazine editors?
Because I'm a scientist myself. Not in the area of atmosphere physics (which, btw, most so-called "climate experts" aren't either), but in the area of statistics and econometrics. And climate change statistics are simply too weak to proof anything.
I'm not saying that climate change isn't man-made. I'd simply want people to not believe everything politicians tell and ask critical questions - for instance, why, if the climate change is man-made, the data doesn't correlate with a positive time-lag if detrended.
I know that this is a rather abstract issue and not easy to understand if you didn't study maths yourself. But consider the following example:
During the past 40 years, McDonald's hamburger sales in the US and and China's economic growth are positively correlated - both have been growing rapidly (even with a positive time-lag, thus this statistical proof is stronger than that of climate scientists). Does this mean that if McDonald's sells more hamburgers in America, China will grow faster? Most likely not. So in statistics, if we detrend the data, we find that the two variables are actually not correlated meaningfully.
This is the very same issue in climate science: we know that temperature abd carbon-dioxide emissions have been raising - but if we detrend them, we can't find a correlation either.
Moreover, even if you find a correlation, you still have to determine its nature: does A cause B or B cause A or does C cause A and B?
That's probably why there's the idiom "never trust statistics if you didn't fake them yourself!"  |
|