Guest Essay by Kip Hansen — 3 March 2024 — 2100 words
Omnia in mensura et numero et pondere disposuisti — Thou hast ordered all things in measure, and number, and weight.
– credited to Solomon’s Book of Wisdom
The basis of physical science is measurement. Measurement is just another word of quantification. Quantification is another word for counting.
In addition to quantification, science entails the qualification of things.
quantify means to find or calculate the quantity or amount of (something).
qualify means to characterize by naming an attribute, basically it means to state any property or characteristic of something. [ reference ]
Taxonomy is a qualitive science – it classifies life forms according to types, characteristics, etc.
The so-called “hard sciences” depend on quantification: measurement and counting. [ reference ]
There is little controversy about the importance of measurement and counting in the enterprise of the sciences, despite the occasional objections from philosophers.
In our modern Mass and Social Media world, numbers are presented to give a sense to “factualness” to ideas. It has been known that numbers have been used to tell lies probably since the beginning of the general use of numbers. “How to Lie with Statistics” by Darrell Huff was published in 1954 to explain this phenomenon and became a classic.
Forbes published an interesting piece by Christopher Kim in 2013 “6 Ways Numbers Can Lie To Us”. Kim’s list includes:
1. Small sample size: Drawing conclusions from small samples
2. Using Big Meaningless Numbers: 14,097,321 . . .
3. Correlation, not causation: Numbers stated in such as way as to imply causation, when they only show a correlation
4. Selection bias: using numbers imply that data came from a random sample when in actuality, the sample has been carefully (or carelessly) chosen
5. Visual trickery: Think Global Temperature graphs with a top-to-bottom range of 3 degrees, to make the increase look huge and alarming or this example:
6. Arbitrary cutoffs: “This is another form of selection bias. Setting arbitrary start-and-end points that impact the meaning of data.”
Great list, but certainly not exhaustive.
And the Biggest Omission? Failing to admit that Numbers are Just Numbers. Numbers are not the things they quantify. Sounds so obvious, doesn’t it? Of course, just telling you the number “687” isn’t useful or informative if I don’t also tell you “687 whats” – 687 apples, 687 inches of string — 687 degrees Celsius – 687 touchdowns.
Similarly, telling you “627 then 687!” has the same problem – nonsensical without the “whats” and “whens”.
And the #2 Biggest Omission? Failing to make a clear statement of exactly what was counted/measured and exactly how the counting/measurement was done. (this could and often does extend to exactly whens.)
In order to prevent the Biggest Omission – How many whats? – every number needs to be accompanied by (even if just implicitly) a clear statement of what has been counted and how it has been counted.
If we want any number to be considered scientific, the rules for this specification [“describing or identifying something precisely”] become stronger and stronger – we should consider this requirement paramount. In a scientific journal paper, this information is sketched out in the ‘Methods’ section and hopefully is more fully specified in the Supplemental Materials.
When this is not done properly, we end up with numbers and graphs like those we see in the Climate Change field: Global Sea Level Rise and Global Surface Temperature. The general public, encouraged by the activists, climate change crisis advocates and complicit journalists, are led to believe that these “numbers” are something real and are, in fact, the thing they are labelled: that the numbers on the graphs are something that could be found in the physical world. This is not true – and I have exhausted myself explaining this here in the past.
Today I will share an example that has created a controversy in the field of medical statistics. An unlikely topic for discussion here at WUWT, but it is a near perfect example and will avoid all the food-fighting over Climate Change.
Maternal Mortality Rate
The topic pops up in the journal Science in an article that is the push-back to another study: “Have U.S. deaths from pregnancy complications tripled? CDC pushes back on study claiming overestimates”
The media has been reporting that maternal deaths [Maternal Mortality Rates] have “spiked”, “climbed dramatically”, “are getting worse”, and that we have an “unacceptably high U.S. maternal mortality rate”.
These stories are reporting the CDC’s announcement, released in March 2023, titled “Maternal Mortality Rates in the United States, 2021”. [ or as a .pdf here ]
The report starts with this:
“A maternal death is defined by the World Health Organization as “the death of a woman while pregnant or within 42 days of termination of pregnancy, irrespective of the duration and the site of the pregnancy, from any cause related to or aggravated by the pregnancy or its management, but not from accidental or incidental causes. Maternal mortality rates, which are the number of maternal deaths per 100,000 live births, are shown in this report by age group and race and Hispanic origin.”
The news carried by the media is based on the simple statement:
“In 2021, 1,205 women died of maternal causes in the United States compared with 861 in 2020 and 754 in 2019 (2). The maternal mortality rate for 2021 was 32.9 deaths per 100,000 live births, compared with a rate of 23.8 in 2020 and 20.1 in 2019.” And this graph:
Clearly, as shown the total U.S. MMR [Maternal Mortality Rate] nearly doubled from 2018 through 2021.
That is a shocking statistic. To round out the picture of MMR, here’s two international views:
Our World In Data supplied the above charts on MMR around the world. The little inset in the left panel shows the slight uptrend in the US MMR over the period reported by the CDC. The good news, that MMRs have dramatically fallen, almost everywhere, to near zero since 1950 (invention of antibitoics, I suggest) is not mentioned. But the barely visible uptick in U.S. MMR is shouted from the rooftops and front pages.
The reported increase in U.S. MMR was so shocking that a group of maternal health researchers, headed by K. S. Joseph, School of Population and Public Health, University of British Columbia, decided to re-evaluate the CDC data. Earlier this month, on 12 March 2024, their paper was published in the American Journal of Obstetrics and Gynecology and titled: ”Maternal mortality in the United States: are the high and rising rates due to changes in obstetrical factors, maternal medical conditions, or maternal mortality surveillance?” [The study is Open Access and available for download as a .pdf from this page ]
Their title gives away their suspicions:
“National Vital Statistics System reports show that maternal mortality rates in the United States have nearly doubled, from 17.4 in 2018 to 32.9 per 100,000 live births in 2021. However, these high and rising rates could reflect issues unrelated to obstetrical factors, such as changes in maternal medical conditions or maternal mortality surveillance (eg, due to introduction of the pregnancy checkbox).”
The big story is this: “But controversy broke out last week over just how bad the situation is, when a paper by academic epidemiologists published in the American Journal of Obstetrics and Gynecology (AJOG) provoked unusual pushback from the U.S. Centers for Disease Control and Prevention (CDC). The paper suggested a widely reported tripling in the U.S. maternal mortality rate (MMR) over the past 2 decades was in fact largely due to a CDC-led recording change on death certificates, the addition of a “pregnancy checkbox.””
“The “pregnancy checkbox” was inserted on death certificates starting in 2003 to address what was at the time a widely acknowledged, substantial underreporting of maternal mortality: At the time, as many as 50% of physicians completing death certificates failed to report that a woman was, or was recently, pregnant. On death certificates, physicians now are asked to check a box indicating a person was pregnant when they died, or within 42 days of the end of the pregnancy. Doctors are not to check the box if a person died of accidental or incidental causes unrelated to pregnancy, for instance, in a car crash or from a gunshot wound. Although the agency rolled out the feature in 2003, it took 14 years before all 50 states adopted the surveillance tool. After that happened in 2017, the agency began to compute the nationwide rate using the checkbox.” [ as reported in Science]
Now we see the issue here. There was a change in what was being counted. When did this change? 2017. When did MMR start “skyrocketing”? At the end of 2017 (years 2018 onward). Once all 50 U.S. states had a checkbox covering possible pregnancy, the CDC started using the checkboxes (counting checkboxes as opposed to counting maternal deaths) to determine Maternity Mortality Rate.
The Joseph et al. study concludes:
“The high and rising rates of maternal mortality in the United States are a consequence of changes in maternal mortality surveillance, with reliance on the pregnancy checkbox leading to an increase in misclassified maternal deaths. Identifying maternal deaths by requiring mention of pregnancy among the multiple causes of death shows lower, stable maternal mortality rates and declines in maternal deaths from direct obstetrical causes.”
The Joseph study looked at death certificates and only counted deaths as Maternal Mortality if the death certificate actually listed pregnancy as one of the contributing causes of death. “Cause of Death” is seldom simple short of something as obvious as a bullet to the head — I covered this during the Covid days in Cause of Death: A Primer – that essay has this image of a death certificate (which also shows the pregnancy checkbox, labelled “If Female”, in the center):
Joseph et al. maintain that the death be counted as a Maternal Mortality only if pregnancy or a related issue of childbirth is specifically mentioned in Parts I or II. When the counting is re-done that way, Joseph et al. found “stable maternal mortality rates and declines in maternal deaths from direct obstetrical causes.”
It is that conclusion that has resulted in a broadside attack on Joseph et al. from the other stakeholders in maternal health, including the CDC itself [quoted] and the American College of Obstetricians and Gynecologists (ACOG). Many mass media outlets covered the story slamming Joseph et al. as threatening to “… reduce the U.S. maternal mortality crisis to an ‘overestimation’ is irresponsible and minimizes the many lives lost and the families that have been deeply affected.” [ source ]
This counting controversy isn’t restricted to the Joseph et al. paper. The CDC’s very own National Center for Health Statistics (NCHS) in a report dated January 30, 2020, had previously reached the exact same conclusions as Joseph: “NCHS found that the increase in maternal mortality in the United States is not likely due to a true increase in the underlying extent of maternal mortality. Rather, the majority of the observed increase in the MMR is attributed to changes in data collection methods (i.e., the gradual adoption of the checkbox). Based on the pre-2003 coding method, the MMR was 8.9 in 2002 and 8.7 in 2018.”
Bottom Lines:
1. Nothing about this controversy changes the real-world number of women who died. Those women died, from whatever cause. Their families, their children, their husbands, their parents suffered their loss.
2. But the over-count has made a big difference in health politics. If Maternal Health stakeholders can point to alarming statistics and create a National Health Crisis from them, more sympathy and money will pour into their cause. More attention and money may actually be a good thing if it leads to more research and actions that can reduce maternal deaths.
3. However, it is never a good thing to create a crisis out of the miscounting, mis-measurement, and mis-reporting of numerical facts.
4. The numbers you see reported (and hyped) in the media are probably false, mis-counted, mis-measured, mis-labelled and do not factually represent the thing they claim to show. [follows from John P. A. Ioannidis, “Why Most Published Research Findings Are False”.]
5. Always, if it is important to you, carefully dig in to find out exactly what was counted/measured and exactly how it was counted or measured.
# # # # #
Author’s Comment:
I shouldn’t have to point out the parallels between this controversy and Climate Science. Alarming numbers are created, labelled as something shocking, public is alarmed and politicians react. Cooler heads reexamine the numbers and point out that “it isn’t really that bad”. Cooler heads are attacked and vilified (even though the official IPCC science agrees with them).
We can all be fooled by numbers – this seems to be a human trait. Personally, I think it stems from a deep innumeracy. This tendency can be overcome with doing due diligence and applying critical thinking skills.
At minimum, we need to ask: What exactly did they actually count? Exactly how did they count it? Does the number really represent the thing (idea, physical fact, actuality) they say it does?
Thanks for reading.
# # # # #
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
That sort of thing is what I suspected, an artifact of reporting or an artifact of a changed definition. I am suspicious, though, of just who changed the reporting protocols, and if it affected their possible future funding.
Tom ==> From CDC statements and reporting by others, the CDC believed, based on its own studies at the turn of the century, that MMR was being vastly under-reported. Their solution was to add the “pregnancy box” to death certificates in all 50 of the United States. That took a decade or so. By 2017 all of the states had added the pregnancy box, and at that point, the CDC started counting “checked checkboxes” as if they were Maternal Deaths.
Failure to post a note that reporting procedures had changed, and the before and after figures were not comparable is probably activism, not health reporting.
Tom ==> worse than that, I’m afraid. The CDC claims that the checkbox counting is finally “correcting the under-count”. Not even their own stats people believe that – that is a statement like those made in IPCC Summary for Policymakers, a political statement.
Are you trying to say that 42 might not be the correct answer?
Very good.
Depends on which universe.
and which question.
You beat me to it.
It is the correct answer as long as we retrospectively invent the correct question.
Don’t panic.
Great post again Kip.
I relate to the “exactly what is being counted” question, as it was my first real “discovery” as a cadet auditor in a large foods processing company.
Wastage of containers (cans) was being reported as the difference between the number of ready-to-ship cans of produce at the end of a production run, as against the number of empty cans signed out of the cans storage silo.
Seemed too simplistic to me.
So I wore out a bit of shoe leather following the usual proceedings with cans supply to the production line, and what happened thereafter.
Turned out that “signed out” from the silo was not “total used” in the production runs at all. Contingency extra was always delivered.
And “left-overs / unused” were returned and signed back into the silo next day.
So that meant 2 new counting protocols had to be instituted, because at both ends of the cycle, “what was being counted” was dead wrong.
How prevalent are these kinds of situations across all levels of human endeavours I wonder?
Mr. ==> Great example. Thanks!
Thanks for the article. I have a hard time explaining these issues to friend and now I can point them to an article with a clear explanation of why and what you have to question when you read statistics or scary numbers.
Trying ==> Hope you find it useful.
I wrote a whole ebook about the general problem Kip discusses, The Arts of Truth. (Even the title isn’t what it seems.)
First chapter discusses the philosophy of ‘truth’ and ‘untruth’.
Next few chapters give many examples broken into different general deception categories—some deliberate, others not (for example Bayes Theorem and the mammogram problem).
Next to last chapter is long, and applies all the previous lessons to climate change. Was reviewed by Dick Lindzen weeks before he retired from MIT. Wonderful day with him on campus.
Last chapter is a wrap with some practical suggestions along the lines of ‘trust but verify—sometimes. Other times do NOT trust (biased sources), only verify. Education examples (biased teachers unions) included classroom size (smaller isn’t better) and standardized testing (scores go up when tests are dumbed down).
Rud ==> Your book is still available via the major eBook outlets, such as here.
Thanks Kip. I preferred my publisher’s iBook version to the Kindle version, because the Apple ebook reader has more functionality than Kindle. Have not bought a paper book for now probably 20 years—ran out of physical shelf space.
Rud ==> You’ll have to supply a link for that version — couldn’t find it in a 1 minute search. 🙂
Kip, I did a search on Apple Books for “Istvan”, which didn’t come up with any of Rud’s works in the first 20 or 30 suggestions. I then tried searching for “Rud Istvan” and got “Gaia’s Limits”, “Blowing Smoke” and “The Arts of Truth”, with each costing $9.99.
Erik ==> Thanks….
“If Maternal Health stakeholders can point to alarming statistics and create a National Health Crisis from them, more sympathy and money will pour into their cause.”
The CO2 concentration in the atmosphere has ALSO been increasing right along side the maternal mortality rate from 2018-2021.
Clearly, global warming is a contributing factor to the alarming increase in maternal deaths. Send more funding IMMEDIATELY to stop the catastrophe!***
***Research funding should be directed to the “Pillage Idiot Luxury Travel and Retirement Fund”.
pillage ==> It might not be that greedy of a motive — doctors and nurses that have dedicated their lives to mothers, babies, and maternal health know that they could do a better job protecting those mothers with more funding and maybe more research would find some way to save some of those mothers.
There are some easy fixes. Grassroots Health in South Carolina tried giving women (mostly black) 2000 IU vitamin D and virtually eliminated pre-eclamsia and dropped preterm births over 50%. This work was done more than 10 years ago and required getting IND permission because the IOM says more than 800 IU is toxic.
https://www.grassrootshealth.net/project/protect-our-children-now/
Not enough of the “concerned” are looking for cheap solutions.
I take 5000 IU of vitamin D almost every day because I am low and that is how much I need to take to bring my D concentration in my blood up to the normal range. Doctor recommended I start at 2000 IU and see if that was enough.
Might there be other causes?
In 2022, 73,654 people died from a fentanyl overdose in the US, more than double the amount of deaths from three years prior in 2019. Fentanyl deaths have increased every year for the past decade, but 2022 marked the smallest year-over-year growth at 4.3%.
How many of the “861 in 2020 and 754 in 2019″, were drug using mothers to be?
But Rush was right again, “Mothers and Children were hardest hit”.
DD More ==> Sorta why Joseph et al. insisted that Parts I and II be used instead of “just” the checkboxes.
Pregnancy/childbirth they felt, had to be at least mentioned as a cause or contributing factor in the death.
Dying of Covid a month after childbirth or abortion is NOT a Maternal death.
Separate general comment related to Kip’s excellent post. FOLLOW THE MONEY.
Look who objected to the changed MMR counting truth—those who would financially benefit from the untruth (CDC, ACOG).
Same is true times near infinity for climate change. Many lucrative academic careers built on basic climate untruth (Mann, Hansen, Viner …). Entire industries (renewables, EVs) built on that untruth plus the additional ‘renewables are grid viable’ untruth.
In an inverse ‘follow the money’ sense, look who isn’t playing along on climate. China and India cannot afford the UK and Germany pretenses. Of course, it is only a matter of when , not if, UK and Germany cannot afford the pretense either.
Actually what was left out was the difference between counting and measuring. Counting is obvious. Measuring is ALWAYS an estimate, not a count. And the laws of significant figures must be applied to measurements. Why?
Because measuring with one instrument that has a resolution of 1 degree, and another instrument that has a resolution of 0.1 degree and then averaging them together will fool most people. I see this in climate reporting daily.
If you average 99 thermometers with the 0.1 resolution with 1 thermometer with 1 degree resolution, no matter WHAT the numbers seem to average out as can be no more precise than the least precise instrument. Therefore, the average of the above may NOT have any decimals. Why?
Because the instrument with 1 degree resolution can not show a half degree difference. That reading is an average, so a 5 degree reading could be anywhere from 4.5 degrees to 5.4 degrees and it will still read 5 degrees.
jshotsky ==> Technically correct, and if you have been following my stuff here at WUWT, you’ll know that I totally agree about temperature measurements and “averaging”.
But I am writing pragmatically. It doesn’t matter if one is counting “1/1000ths” or “1/10ths” or apples or Maternal Deaths for my point to be made.
There is a difference between counting things – people, cans, apples, dollars – and measuring with an instrument. Counting objects results in no uncertainty – you can always count again to verify if in doubt. There should never be any uncertainty about the number of dollars in your checking account. Measuring always results an some level of uncertainty due to typically several sources like calibration uncertainty, resolution etc. That is why standard practice requires including a statement of measurement uncertainty when reporting results.
Note, however, that we often count things in statistical sampling processes such as counting defective widgets in a sample from a production run and then use math to estimate quality of the entire lot. That brings uncertainty back into the process and requires appropriate determination and reporting of variance at specified confidence levels.
Rick C ==> There are differences between simple counting and scientific measurement, sure and of course. And I have written endlessly about how measurements have to come with a correctly stated uncertainty.
But for thepoint of this essay, measurement is just another kind of counting.
“Counting objects results in no uncertainty”
This isn’t really true. There can be lots of uncertainty resulting from how you count, and the definition of what you are counting. That’s the point of this article.
Bellman ==> Again, I have to agree with you…counting apples…by the pound, by the peck, by the bushel, by the pallet-load. All valid measures/counts for apples. Even a one-by-one count can require a stated uncertainty, depending on the counting methods….imagine a nine-year-old counting ten thousand apples….what’s the uncertainty?
And, quite right — this article is about why it really really matters: exactly what and exactly how.
“If you average 99 thermometers with the 0.1 resolution with 1 thermometer with 1 degree resolution, no matter WHAT the numbers seem to average out as can be no more precise than the least precise instrument.”
Except your 1 thermometer with a 1 degree resolution will only contribute 1/100th to the average.
Take a 100 random values, rounded to 1 decimal place, take the average, then round one of the figures to the nearest integer. How much of a difference will that rounding make? At most you’ve changed that single value by 0.5, so at most the average can change by 0.005.
Nope. You can make the mathematically derived average do exactly that but it won’t make any difference – you have to take into account what is being measured. These are not abstract numbers, no thermometer is capable of measuring to 0.005 so your mathematically derived statistic is completely meaningless in this case. The correct answer is as stated above, “no more precise than the least precise instrument.”
bellcurveman will never acknowledge the truth.
“These are not abstract numbers, no thermometer is capable of measuring to 0.005…”
You’re missing my point. It’s not about the precision of the thermometer, but about how much that one low precision thermometer can change the precision of the average.
I’m not missing your point at all but is very clear that you missed the point of the post and all the comments below that. If you are just dealing with abstract numbers then your approach would be correct but the example was specifically about measurements in the real world. With real world measurements you have to take into account what is being measured (and its specific properties) and how it is being measured – your comments completely ignore both of those factors.
I’m not commenting on this article, just on one specific claim made in a comment.
Real world measurements are made with abstract numbers. The maths in this situation don’t change just because you are measuring temperature. If you are taking an average if 100 measurements then an uncertainty of 1°C in just one of those measurements can only affect the average by 0.01°C.
It’s telling that whenever I question these so called laws, or people’s interpretation of them, I’m never offered any evidence to support these claims, just appeals to tradition and authority, and lots of name calling.
No. Real world measurements are not abstract numbers and to treat them as such is very wrong. What if the 99 measured temperatures with 0.1° resolution were slightly above or below the average but the one with 1° resolution was exactly on that average?
People who treat real world data as abstract numbers should never be let loose near real world data, it’d be like giving a box of matches to a toddler.
“Real world measurements are not abstract numbers”
They are exactly that.
“What if the 99 measured temperatures with 0.1° resolution were slightly above or below the average but the one with 1° resolution was exactly on that average?”
Wut? You’re sundowning. Are you discussing systemic error in 99 thermometers and no systemic error in the one? Yes, if all of these 99 thermometers had that, then the average would converge upon a value incorrect by that 99% of that error. Or are these 99 systemic errors distributed? If so, how?
So, yes to get an exact expected value and standard deviation, you need that info. But:
jschlotsky did not aks for that to be included.
Without treatment of systemic error, Bellman’s evaluation is correct.
The larger ANY source of systemic error is, then the more likely it is that it will be spotted and corrected for. Even if circular logicians like you then clutch pearls about “adjustments”
Your hand wave about mystical sources of error that are not mentioned, but that are still large enough to qualitatively change results, are irrelevant.
Hold off on the irony, please.
“Real world measurements are not abstract numbers and to treat them as such is very wrong.”
The point about abstract numbers is they can be applied to many different concrete situations. The rules are the same in all cases. 1 divided by 100 is 0.01 in all cases. If you want to invent a new type of number that only applies to real world measurements – then it’s up to you to define how they work.
“What if the 99 measured temperatures with 0.1° resolution were slightly above or below the average but the one with 1° resolution was exactly on that average?”
Then it would be a remarkable coincidence, and is still irrelevant to the so called laws of significant figures.
“People who treat real world data as abstract numbers should never be let loose near real world data, it’d be like giving a box of matches to a toddler.”
And yet you still can’t provide a single piece of evidence to support your argument. Just endless whining about abstract numbers.
Let me give you a non-abstract example. Say you are measuring the temperature a a substance. You use 99 different well-calibrated thermometers with a digital display given 1 decimal place. Every one reads 17.6°C. The average of these 99 is 17.6°C. Given the resolution you can say that the temperature is likely to be between 17.55 and 17.65°C.
Now for some reason you decide to use another well calibrated thermometer that only gives the reading to a whole number. It says 18°C, which is expected and consistent with all the other readings. You average all 100 thermometers and get an average of 17.604°C. You can write this as 17.6 °C, or you can do what you want and write it as 18°C. If you have your way you are claiming you only know the temperature is between 17.5 and 18.5°C. In what way does your “real world” use of measurements result in a better or more realistic result?
“If you are taking an average if 100 measurements then an uncertainty of 1°C in just one of those measurements can only affect the average by 0.01°C.”
For clarity, you now have 99 of God’s own thermometers, and one with your referenced uncertainty. Right? If so, then right again…
I bet you actually think it is true that an average golfer can score a 4.5 on an easy par 5 hole just because 4.5 is the average for the last 100 golfers to play that hole.
Them you will lose your bet.
Because the instrument with 1 degree resolution can not show a half degree difference.
Any thermometer can be said to show whatever the government bureaucrat who owns it wants the general public to hear … which may have been decided BEFORE the thermometer was observed. Their motto is: Accuracy, Precision and Margins of Error Are for Losers
You can both be correct about your disdain for government bureaucracy, and with the Bizarro world statistical constructs of the Gormans, metoo’ed by km.
Read Bellman’s last post. What he said is true, and so is his inference that 99 similar thermometers would depress the standard deviation of the average of them by a factor of ~10, over just one of them.
And these guys tear their clothing and wail about being ignored above ground….
Read what I posted, reread the post and then, perhaps, you’ll realise that Bellman ignored all of it, along with the information in the example, and treated measurements as if they were abstract numbers. That is why he is wrong and so are you.
I not only read – and now reread – the post, but did the experiment myself. I got Bellman’s exact difference between standard deviations of the average using the 99 thermometers v using the 99 + the one with more error. And of course also the original reduction in average standard deviation between 1 thermometer and 99 that he inferred. Both using statistics 101 concepts, and by bootstrapping. But since understanding the inference is necessary to understanding his 0.005 value, he’s guilty of giving you too much credit.
Both bad wiring here specifically, and laziness with exploring any results that are counterintuitive upon first glance….
Nope – see you’ve just done (again) what I said about Bellman. You both treated the data as abstract numbers and did a quick statistical calculation as if they were just that, abstract numbers with no basis in the real world. Real world data is not a collection of abstract numbers, they are a measurement of something with inherent properties that you cannot simply ignore as irrelevant or inconvenient – that’s very lazy and stupid thinking. The properties and methods of measuring those properties are part of the data and should always be considered as part of the data. If you want to play around with abstract numbers, that’s fine – go right ahead, just stay well away from actual data.
There are many ways to lie with data but the primary climate hoax does not involve any data
The primary hoax is a data free prediction of CAGW
CAGW has never happened so there are no historical CAGW data
There are never data for the future
So CAGW predictions are not based on any data
They are climate astrology
Even worse, CAGW predictions since 1979 have been wrong for 44 years in a row
There is government manipulation of temperature and sea level data, but historical trends, even with the “cheating”, are NOT used for the scary CAGW predictions.
If the historical GAT and sea level data were perfectly accurate, we would still hear CAGW scaremongering. In fact, the government bureaucrat scientists would have had more credibility today if they had always been honest and accurate with historical GAT and sea level rise data.
CAGW predictions would still be nonsense, but honest historical climate data could have created the (false) impression that government bureaucrat scientists had integrity.
The GAT can never be ‘accurate’ – there is no possible way of taking precise measurements over the entire surface of the planet and, even if you could, the different regional climates would render them meaningless. They are a political tool of an overly politicised group of academic activists styling themselves ‘climate scientists’ for no good reason following a political idealogy and agenda.
The precise answer is that GAST is meaningless. The reason is that temperature is an intensive rather than extensive physical quantity, so can never be meaningfully averaged mathematically.
Can it be done—yes. Does it have any physical meaning—no.
Rud ==> You and I the ONLY ones that understand this? eGads!
There are others…
Kip, it’s OK. Many of us asked that question long ago.
Essex, McKitrick, and Andresen wrote a good explanation about it here:
https://www.fys.ku.dk/~andresen/BAhome/ownpapers/globalTexist.pdf
Thanks Kip.
Maybe someone can get this post to Sen John Kennedy or his staff.
It might be useful if an expert witness (from the CDC) tries to use buggy numbers at a committee hearing.
John H ==> Usually doesn’t do any good — they don’t want to hear true facts and accurate numbers.
Sen Kennedy has a reputation of knowing facts and numbers better than the “experts” that are brought in to inform him. The latest being the young skier — really-really bad choice by someone. Prior one was Deputy Energy Secretary David Turk. I find it hard to believe Turk got the job because of his smarts.
It is difficult to engage with the folks at the highest political levels.
John H ==> The pretty boy cross-country skier just want to complain about lower snow cover seasons in some places. didn’t know anything about climate change at all. But he was pretty! An Olympic athlete.
Thought CO2 made up a big part of the atmosphere….
He was looking for a new activism cause to replace his BOLM and defund the police activism. Obviously just seeking notoriety. Did not know what CO2 is.
And we are told via the Global Warming Potential numbers that methane is 86 times more powerful than CO2 at trapping heat, N2O is 300 times more powerful and CFLs are several thousand times more powerful.
What’s being counted is the inverse relation of the concentration of each gas in the atmosphere. The less gas in the atmosphere the bigger the GWP number. The infrared absorption spectrum of each gas has nothing to do with it
@ur momisugly Kip Hansen
The Graph “Figure 1” shows the “Total” as less than the “Non-Hispanic Black”. How can that be???
… Answers own question: because they are mortality *rates* not absolute numbers. Doh!
Government departments and agencies want to employ more people by having them check more and more boxes. Then they count the boxes and write reports. This is not new. It is called “leadership”.
Unfortunately, leadership forgets to bring the counted boxes to congressional committee hearings.
What is new is more and more pages that are left purposely blank except for the sentence that tells you they were left purposely blank. I think we need to count those too.
Well, no, the right technique now is pages that are entirely BLACK. Even the page NUMBERS are “redacted for national security.”
Actually, Sam Clemons had this down back in the early 20th century when he quipped, “There are lies, damned lies, and statistics.” (he attributed the remark to Benjamin Disraeli).
Any argument can be bolstered by reference to statistics if the statistics are falsified, or present less than the wholly characterized universe of descriptive data. You don’t even need to understand math or science to intuitively understand that one needs to “hear the other side of the story” to correctly judge anything in life. It is no more complicated than that.
The decades old running battle between the warmunists and the climate skeptics boils down to the claims by warmunists that “there IS no other side to the story” so listen only to their version of truth.
L Frank Baum had this figured out when he wrote “The Wizard of Oz” in 1900, writing the famous line, “Ignore the man behind the curtain!” The Oz story was not just a fanciful children’s tale but a warning to the world of the power of misrepresentation of facts by people in power who are not who and what they claim to be. He was writing about propagandists in government, of course. But the people failed to understand, and as a result the 20th century became the bloodiest of centuries, all due to populations who accepted the propagandists’ lies.
Duane ==> I never got over the flying monkeys….in the movie version.
Happens all the time in climate change.
Pielke and Richie published ‘How Climate Scenarios Lost Touch With Reality’ in 2021.
“According to Google Scholar, from the beginning of 2020 until mid June 2021 authors published more than 8500 papers using the implausible baseline scenarios, of which about 7200 used RCP 8.5 and nearly 1500 use SSP5-8.5. Neither IPCC nor the broader climate modelling community has sought to counter or reverse this proliferating source of error in projections of future climte change”
“The emerging market for climate scenario products has led to a $40bn ‘climate intelligence’ industry involving companies such as Swiss Re and Mckinsey and others…..These companies are using implausible RCP scenarios to develop predictive products that they sell to governments and industry, who will depend on these products to help guide policy and business decisions in the future”
I think what are being counted in the Climatoverse are algorithms which add up to 1.5C.
“Six Ways Numbers Can Lie to Us”
Four of those ways are well illustrated by all those Pause articles published here.
Bellman ==> Where ever you see these tricks — that’s what they are. Making simple observations of things — such as a multi-year flat spot in a graph, is not the same — it would be if one showed ONLY the flat spot though.
Catch ’em when you can.
One of my bug bears is the use of parametric statistics when distributions are so far from normal that they give invalid results. In one paper I reviewed, “an anonymous reviewer” was thanked for the help, but the usual response of authors is to double down.
Climate science cares about only one parameter: the almighty mean. Everything else is just ignored.
Some of us can easily count to 20. Some of us can easily count to 21.
Some others are just confused.
Gunga ==> There is a very small island off the coast of a larger island, St. Vincent (in the Caribbean) where a high percentage of the people can count to 21 on their fingers…..many have six fingers on one or the other hands.
Then some of us can count to 22? 😎
Long way before we can count to 42!
Gunga ==> That’d be a lot of fingers….
Tell Monk.
😎
Oh that’s nothing. The ancient Babylonians could count to 60 just using their fingers – 1-12 on one hand, multiples of 12 on the other, using parts of the fingers between the joints I believe. It was so successful we still use base 60 today for time (seconds, minutes and 2 12 hour periods for days) as well as degrees for navigation.
A few joints might help make CliSy seem like it makes sense. 😎
PS Thanks for the Babylonian info. I didn’t know that.
Thanks Kip, this reminds me of a an experience I had with a program about Infant Mortality several years ago. As part of my job we held a lot of educational industry events. Given my role I was assigned to provide execution and marketing support to a program for another part of our “large” organization.
For this program, there was an NGO supplying the topic and headlines, which highlighted the alarmingly “high” IM rates in the USA vs other countries as well as disparity of rates among races within the USA. Not being a topic we would cover in my area I read the reports they had attached to the program as well as doing some research on my own.
Although I don’t remember all of the program details, I do remember easily finding in my research that other countries measured IM rates differently. Therefore, they were comparing apples and oranges making the program headlines misleading in that respect.
I also found issues with the reports the the organizer sent me and the program write up. Interestingly, I was contacted by a someone( it may have been a doctor) who asked if we were going to cover the problem of high rates among hispanic americans as that was a concern of hers. The writeup on the program mentioned the disparity of rates between black and white people specifically. However reading the reports attached to the program provided by the organization, the rates among hispanic mothers showed better rates than those of white mothers. When I questioned the organizer on the mismatch of the promotional materials and research reports, the lines went quiet. I saw a big problem being finding the right soliton. Where they were focusing on socio-economic issue, perhaps that date point would move them into a different direction. More info was needed, but I saw that the ‘objective’ of this particular agency fit the narrative they were selling. Sad as there was a real problem there.
But 9 out of 10 readers believe it is truly a crisis.
Scientific notation dictates that the calculation resolution is not deeper that the least significant factor. Just because a computer can add 20 zeros, does not make a measured 1 become 1.00000000000000000000.
Part of the fallacy of numbers is a result of computer resolution.
2020 was in the era of SARS2 virus causing people to contract the disease COVID-2.
A vaccine was available at the end of 2020, but distribution in third-world societies was slow to get going. So 2021 statistics probably include deaths from COVID-19, even in societies like Canada.
(Statistics on cause of death were imprecise, to be charitable. Many deaths were recorded as cardiovascular or respitatory but those organs were badly weakened by COVID-19. (It is a respiratory disease with a substantial cardiovascular impact.
Scientists have to ‘control for’ unusual factors.
I.E. Remove samples with another cause, or at least indicate what was not controlled for.
A great example of the latter is the Limitations list in ‘Association of Cannabis Use With Cardiovascular Outcomes Among US Adults’ in Journal of the AHA. Item seven says “…the large proportion of users being young confounds this study in an important way.”