Mike Lorrey writes- PAY ATTENTION CLIMATE ALARMISTS:
“The phrase ‘correlation does not imply causation’ goes back to 1880 (according to Google Books). However, use of the phrase took off in the 1990s and 2000s, and is becoming a quick way to short-circuit certain kinds of arguments.
In the late 19th century, British statistician Karl Pearson introduced a powerful idea in math: that a relationship between two variables could be characterized according to its strength and expressed in numbers. An exciting concept, but it raised a new issue: how to interpret the data in a way that is helpful, rather than misleading. When we mistake correlation for causation, we find a cause that isn’t there, which is a problem. However, as science grows more powerful and government more technocratic, the stakes of correlation — of counterfeit relationships and bogus findings — grow larger.”
From Slashdot: The History of ‘Correlation Does Not Imply Causation’
==============================================================
From the Slate article referenced by Slashdot:
The graph below, again from Google Books, shows the shift in language that marked this change in spirit: Up until the early 1900s, causation showed up more often than correlation in the corpus; then the concepts flip. (I’ll let someone else explain why correlations have been trending downward since 1976.)
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
wrong again.. .. try..
“rabid warmist activist manipulated urban land temperature”
tomt~ quick question in regards to your contention of rapid warming:
How far did the Earth’s temp drop, from the height of the MWP, to the bottom of the LIA?
How much of that has been regained since the bottom of the LIA?
Thanks in advance.
One accusation that could be levelled at the mentality of the hockeystickers, is that they are not very good at imagining alternative explanations to account for observations. Eliminating incorrect hypotheses is, in fact, the bedrock of the scientific method.
A good scientist is distinguished from a mediocre one by their ability to imagine causal alternatives, and by their diligence in examining them.
Other than that, I agree with David Hoffer’s comments above.
Also, perhaps the nature of books has changed over time in terms of the technical/non-technical ratio. Does the search look for occurrences of the word or just the number of books that use it one or more times? Would this be affected by changes in book length? Perhaps Google hasn’t finished digitising books and is doing them by category, not chronological order. Perhaps they don’t have access to more recent technical publications for commercial reasons. Perhaps certain types of book are now translated into English more often.
I dunno. But the list goes on, and the number of possible alternative hypotheses that could be tested is effectively infinite. The best scientists will break into a smile when they hear the phrase “One experiment too many!”
richardscourtney says:
October 3, 2012 at 1:26 am
Any two time-series will correlate for some periods of the time if they are long enough.
Which is chance. Note I said ‘absent chance’.
Philip Bradley says:
October 2, 2012 at 10:29 pm
Possibly true but not when using only two variables. Not only is “correlation between A and B isn’t proof that A causes B” true but two correlated variable could have a common cause. The correlation between preachers’ salaries in New England and the price of rum have inflation as one of the drivers. Two correlated variables could also be mutual causes: chickens and eggs, so to speak.
http://www.amazon.com/Causality-Reasoning-Inference-Judea-Pearl/dp/052189560X/ref=sr_1_1?ie=UTF8&qid=1349259646&sr=8-1&keywords=causality
Better slogan: DON’T USE STATISTICS.
Stats remove the time dimension, which removes the one thing that can guide you toward finding a causation or connection. If you use graphs and waveforms, paying attention to regularities and PHASE, you’ll always be able to spot the leading and lagging variables… or determine quickly that there’s no connection at all between X and Y.
You’ll note from above that temporally churches can be demonstrated to cause pubs to a high degree of confidence (I forget exactly the r^2 but it was 35 years ago). Post hoc ergo propter hoc, so to speak.
Philip Bradley:
At October 3, 2012 at 3:18 am in response to my rebuttal at October 3, 2012 at 1:26 am of
your fallacious claim at October 2, 2012 at 10:29 pm saying
you quote from my rebuttal
and reply
You cannot pretend that ‘chance’ may be “absent”, and that is why your assertion is plain wrong.
Richard
polistra says:
October 3, 2012 at 3:35 am
I guess you never heard of autocorrelation of time series.
If the time variable has been removed in an analysis, it was because it wasn’t considered or wasn’t deemed important. Statistics says nothing about whether time should or should not be a variable.
Doesn’t work very well for chickens and eggs. The actual beginning is the only place where the first of two correlated cyclical events can be determined OR with an experiment which establishes a beginning.
I like Go Home’s correlatons. Great.
ThomasT – Logic, does not seem to be your strong suit.
As you head back to the safety of your Church of CAGW remember that those hoofbeats behind you are probably not from zebras, horses, or bison. The are from the sensible, scientific bloggers on WUWT.
Open your mind please. The Alarmists have taken a short term correlation of global temperatures and rising levels of CO2, and extrapolated them with computer models that made forecasts/predictions that never came true.
theduke says
experienced over the past 100 years because establishment climate scientists refuse to vigorously explore (and disprove) other possible causes should be offensive to anyone who respects the scientific method.
———-
I’m pretty sure you have not vigorously explored the scientific literature to make this claim stick. And insisting on some arbitrarily high level of proof would be cheating.
D Boehm says
D Böehm on October 2, 2012 at 9:23 pm
Thomas T,
Positive feedbacks have been falsified by Planet Earth; the ultimate Authority. Believe what you want, it isn’t reality. CO2=CAGW is a debunked fantasy.
————–
Simply saying that positive feedbacks don’t exist isn’t good enough. Putting words in Natures mouth isnt a convincing argument either. It’s very likely you don’t understand positive feedbacks and how they apply to climate.
See what happens when you graph “global warming” and “anthropogenic climate change”. I’ve done it for Google Books and Google Scholar, the former stimulated by this post, the latter something I did back in February.
richardscourtney says:
October 3, 2012 at 3:45 am
You cannot pretend that ‘chance’ may be “absent”, and that is why your assertion is plain wrong.
Chance for the purposes of this discussion is a function of sampling. It is not a property of the real world.
Chance is absent from the real world.
Thomas T says:
October 2, 2012 at 8:53 pm
What is important here is that at present no other parameter, such as solar or earth tilt, etc, correlates well with the observed rapid increase in global temperature.
—
That would be important, if it were true.
The secondary factor that you are employing to improve your case is another phrase that I’m sure has also increased in recent decades. It’s called Cherry Picking. That is, pick the years where your hypothesis holds and discard all the others.
In this case if you extend your timeline back a few centuries, and your forward to today, you find out that there is no correlation between CO2 levels and temperatures.
“(I’ll let someone else explain why correlations have been trending downward since 1976.)”
The answer to this is fairly simple. With the information/computer/Internet revolution, we have seen an insane explosion in available information. Frankly, humans are not yet capable of dealing with the information overload. The reduced use of the word “correlation” reflects our inability to relate these large amount of diverse data. It is, in other words, an indicator of our psychological state of mind, in which uncertainty is growing and making sense of the world is increasingly challenging. This trend also explains why people have become increasingly ideological–they cling to simplifications (not guns and religion!) to make the world more manageable.
Dale says:
October 3, 2012 at 2:15 am
I get older by 1 per year. CO2 goes up by 2 ppm per year.
Therefore my age determines CO2 ppm.
—
You have cause and effect reversed.
Increasing CO2 is causing you to get older. If we could start lowering CO2 levels you would find yourself getting younger.
Philip Bradley says:
October 3, 2012 at 3:18 am
Which is chance. Note I said ‘absent chance’.
—
Which is of course the kicker.
It is impossible to prove whether your correlation is real, or the result of chance.
Philip Bradley says:
October 3, 2012 at 5:37 am
Chance is absent from the real world.
—
This has got to be one of the stupidest things I have ever read.
Now the burning question is- do ice-cream sales affect swimming pool drownings or vice-versa?
If eating ice cream were known to cause, say, stomach cramps that increased the likelihood of drowning, then correlation would be intuitive toward causation. But there is no intuitive link, and so your example is weak.
Whereas, for over a century CO2 has been known to be a greenhouse gas.
Could some one comment on something I read on another blog? (I have no science background other than high school)
They claimed that co2 had an absorption rate of around 8% and a concentration of about 380ppm in the atmosphere. While H2o (water vapor) has an absorption rate of over 90% and a concentration of 30,000ppm. Is this true? So how could co2 be anything but completely negligible regarding climate change?
michael hart,
One accusation that could be levelled at the mentality of the hockeystickers, is that they are not very good at imagining alternative explanations to account for observations. Eliminating incorrect hypotheses is, in fact, the bedrock of the scientific method.
Sure, but suppose we have a hypothesis which is based on well understood physical mechanisms, which can’t be eliminated, is consistent with our observations and indeed predicted them in advance. Suppose the obvious alternative hypotheses have been examined and can’t account for observations. Suppose no one has so far put forward an alternative hypothesis which is actually supported by hard evidence. How much time, effort and resources are we supposed to devote to looking for alternative explanations which are unknown in nature, whose existence is entirely speculative, and which are certainly not required to exist? And if we at the point where we are looking for “unknown unknowns” then how will we ever know we have eliminated all possible alternatives? Surely there comes a point where we have to accept that we have an explanation which works well enough to be at least provisionally accepted?
Now I’m not suggesting that we stop trying to improve our knowledge of the way our climate works, people are indeed still doing this and no doubt there are still gaps in our knowledge which will hopefully become smaller with time. But there is no particular reason to assume that this will lessen rather than increase the case for recent warming being caused by human GHG emissions.
The correlation coefficient between the series 1,2,3,….10 and 100.1, 100.2, 100.3,….101 is 1 (perfect correlation). The problem is that with two linear series, perfect correlation means absolutely nothing.
The classic example of spurious correlation was provided by George Udny Yule, a Past President of the Royal Society and a colaborator of Karl Pearson in the development of statistics.
A study of Church of England marriages betrween the years 1866 and 1911 showed a coefficient of correlation of +.95 with standard deviation of only .014 when measured against the standardised mortality of 1,000 in England and Wales. Thus providing clear evidence that marrying in church increased the risk of death.
Actually, as Udny Yule pointed out, both series were reducing over the whole 45 year period and the only signifcant correlation was between each series and time.
While there is evidence that human activity contributes to warming (e.g. UHI) and there is also evidence that human activity adds to CO2 levels, the correlation between any increase in CO2 levels and warming may well be spurious. (I am aware of 150-year old laboratory experiments but I am also aware that what works in a lab experiment more often than not does not translate well inot nature.)
Both series (if measurable) are correlated directly to increasing numbers in the world population and to the increasing average wealth of that world population. Even if all energy were to be provided by “renewables” human activity would still contribute to warming so long as either or both the numbers and wealth of the world population continued to increase.
Alberta Slim says:
October 3, 2012 at 4:05 am
ThomasT – Logic, does not seem to be your strong suit.
As you head back to the safety of your Church of CAGW remember that those hoofbeats behind you are probably not from zebras, horses, or bison. The are from the sensible, scientific bloggers on WUWT.
Open your mind please. The Alarmists have taken a short term correlation of global temperatures and rising levels of CO2, and extrapolated them with computer models that made forecasts/predictions that never came true.
Repeated for effect.
* * *
Andrew Adams,
You have described the climate null hypothesis, which has never been falsified. Nothing observed today is unprecedented or unusual. It has all happened before, and to a greater degree, during times when CO2 was much lower.
Therefore, the null hypothesis remains unfalsified, and the alternative hypothesis of CO2=AGW remains an unproven conjecture with no empirical evidence proving that it exists.
Graph against temperatures please.