/sci/ - Science & Math

>>8566497
>Look at my picture!!!
>Look, I stapled high frequency, high variance instrumental data onto low frequency, low variance proxy data. So climate change is true!

A statistical abomination. What happens if you consistently use the same data type (non-instrumental) and don't conveniently redate proxies (unlike Marcott)? You get pic related. No big deal. And no hockey stick.

>>7909495
Yeah, like the Medieval Warm period didn't exist. Whatever.

>>7750011
Here's an improved version.

>>7526436
>gradual increase in temperature
> Anthropogenic Climate Change

Its been warming since the Little Ice Age. Did they have SUVs in 1750?

>>7290790
And guess what. Here's a temperature graphic that's made without sewing instrumental data to the end. No hockey stick!

>>7057304
Well lets look at a much larger picture to get better context. Looks like nothing significant is going on.

>>7051415
Oh boy, a silly hockey stick graph. Take low variance data, low correlated data (that still represent "global temps") "proxies" and sew high variance data onto the end (usually instrumental data, e.g., thermometers). At the least, this is horrible statistical technique. Quite possibly deliberately deceptive.

See attached. A construction where proxies are not mixed with instrumental data, creating statistical meaning instead of statistical garbage.