Posted on 11/22/2011 6:05:34 AM PST by reaganaut1
PhD Statisticians are well aware of all the abuses but cannot control what the world does with numbers and data. That is why it is important to use a statistician with a good reputation.
Many false positives and false negatives are the result of poor statistical design.
Just as guns don’t kill people but rather people kill others using guns, statistics don’t lie so much as people lie with statistics.
This is a hit piece on people called ‘statisticians’ when in fact such people are not statisticians but rather are abusers of statistics. Such abusers should be distinguished from trained, experienced and objective statisticians.
A trained, experienced and objective statistician will most often look at the data sampled and collected by others and call out the biases in the sample and report that the data or information is inconclusive or non-suggestive of anything.
Abusers of statistics will fish around until they find a false correlation of some sort which is easy to do and then report it sometimes knowing full well that it’s bunk.
In this context, statistics is very similar to political grandstanding or to a lawyer’s brief where they present facts that only favor their client.
The trained, experienced and objective statistician is like the Judge on the bench that reads the facts and applies the law in balance to lead to optimal justice. The abusers of statistics are like lawyers that are biased in their presentations and will not have neither facts nor law for their arguments but will resort to hyperbole and character smearing.
Abusers of statistics will nearly always report something because they will not allow themselves to report nothing even when there is truly nothing to be inferred from data.
Trained and credible statisticians will most often report nothing because most data is poorly collected or collected according to a flawed sampling design that has embedded selection bias.
“This is news? The Global Warming crowd have been running this con for years.”
Not really. They don’t even bother trying basic statistics on the projection models. To do that, you would have to calculate standard error in model predictions. The AGW guys rarely do that.
Ah, but can the two scientists making these claims prove them?
In effect turning the weapons of statistical analysis against their own side, the trio managed to to prove something demonstrably false, and thereby cast a wide shadow of doubt on any researcher who claims his findings are statistically significant.
Wait a second - aren't they committing the very same error they claim to refute; namely, using statistical analysis to prove something they claim to be statistically significant? I dunno. The whole study seems to me to be self-vitiating.
Cordially,
Cordially,
Obviously you missed that episode of Mythbusters....
From Foreward of Huff's book cited in Post #2, author noted there is Disraeli.
One of the best books you’ll never find in a government school.
Don’t wonder. It’s 100%.
Wow! That is one shiny turd!
It’s a really nice color too.
Research shows that 82.535% of all statistics are simply made up.
“You cant polish a turd, but you can roll it in glitter!”
Timeless wisdom.
It is up there with, “Whether the water is salt or fresh, sh!t floats.”
Thank you.
This class is how to lie with figures and make figures lie, and the graph is you biggest ally in creating the lie.
Anyone that uses the word statistics in this class will get an “F”!
Interesting points, and thanks for the post. I believe, however, that you are in an endless loop of non-provability here.
Not one statement of fact in your post is provable without scientific research backing it up, and all that research is suspect unless, indeed, the statisticians used can “be distinguished from trained, experienced and objective statisticians.”
This distinction in general would require scientific proof backing up its validity, and specifically, for instance, an “objective” statistician would need provable criteria by which to define him though examination of data. Researchers would collect that data, and then a statistician would be needed to evaluate that data to make sure those criteria were met, but that second statistician would also need to be proven objective. And so on.
In addition, reliance on reputation is the antithesis of the scientific method. Indeed, it was the authority of Aristotle’s reputation - which reigned in academia throughout the Middle Ages - that held science back until it was dethroned by the scientific method, ushered in by the philosophies of Descartes and Bacon.
I believe - but of course can not prove - that between the slipperiness of language, and the corruptibility and fallibility of human nature, we are arrogant in assuming we know very much of anything beyond the obvious, and we usually see the obvious through quite imperfect lenses as well.
Ah, but you can polish a turd. Mythbusters proved it to be so. It is however, still a turd.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.