« Harassed Owl | Main | In my mind, Shirley Sherrod is owed an apology »

July 16, 2010

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

jmotorhead

Having taken an advanced statistics class in college you also, need to know how they calculated because that can effect the outcome as well. As my teacher said, statistics do not lie, however, liar's use statistics. You need to know the difference.

Brian R

Not the first time I've heard something like this. There was an even more egregious example in Australia, where the only station in a 500km radius was "adjusted" to match the trend of the other stations. Yeah, totally makes sense.

OTOH, there are sites that claim the urban heat island is bogus.

Personally, I've given up. Basically everyone is claiming that the other side is lying, and that the other side is lying about our side lying, rinse/repeat ad nauseum. I have no way to disentangle it short of evaluating the data myself, which I'm not qualified to do.

J. Carmichael

Dave,

I don't entirely disagree with you, except with one point, (while right now you and I may be debating the color of a rain drop) I'll concede that pinpointing a location on this small of a map is an exercise in extremes of standard deviation, but on this particular image (once clicked on) I am of the inkling that the "X" is nearer the 4-6 delineation than the 0-2 band. which based on your pretense would make this "Just Wrong" ;-)

Either way, I would like the scientific community to find a set of accepted data points... analyzed using the scientific method, to use for all reports. I'm inclined to agree with many scientists (not even those aligned with the "Skeptic" community) that also call into question using NOAA reporting stations that are in urban areas where extremely acute conditions can affect the data erroneously. (Seriously some are at the departure ends of Jet runways!). I do think using the satellite data is a more fair average.

Dave

Spin is always entertaining.

Note that the "X" is in an area that, according to the chart, is in the 2-4 degree range. To make his point, the author calls the this "4 degree contour" and therefore claims that the data is off by 3 degrees. But it's clear that the labels are on the edges of the color bands (thus each color represents a range). So in reality, it's off by only 1-3 degrees. And frankly, that "X" isn't all that for away from the 0-2 degree band, which would be just right.

No approximate curve is going to accurately represent every point.

The comments to this entry are closed.