What Are They Really Counting?

Guest Essay by Kip Hansen


WARNING: This is not a technical essay. There is almost no science in it. It is not about AGW or any issue involved in the Climate Wars. My editor describes it as “chatty”. It does ask two extremely important questions.

We are all constantly bombarded by numbers….in the press, on the radio news, on the TV news, here at WUWT. Numbers as sheer numbers, numbers as graphs, charts, images, and in words and more words. Putting a number with an idea has a magical power over our minds – it makes the idea ‘more true’ – it offers to our minds a sort of proof for ideas and concepts.

In this essay, I look at an important question, one we must all ask – ask ourselves and ask the sources of these numbers – What exactly are they really counting? In our little introductory image (cute, huh?) we see they are counting  “counting bears”. 1 bear, two bears. But, what exactly? In the upper panels, they are counting green plastic counting bears. Their count = 1 (and in words – one). In the bottom panels, they are still counting plastic counting bears, but one red bear and one blue bear, or 2 (in words – two) bears altogether. Even more exactly though, the bottom panels have one red bear, one blue bear and zero green bears.

This is not just being fussy. When we have only the information in the top panels, we count one bear (and maybe note that it is green). As far as we know, all bears in this context are the same color, and color doesn’t matter. If these were real seal-eating/fish-eating bears, some might be white, some brown, some grizzled, some a cross-breed mixture. For biologists, the difference is important – refer to the Polar Bear Wars. For a camper on the tundra, one very hungry bear, possibly man-eating, is more than enough, regardless of color.

For this kindergarten example, we see that even in the most elementary types of counting , there are details that may need to be explored and explained.

Just to be clear, all measurement is the same as counting in this regard.

meas·ure   ‘meZHər/   verb      to ascertain the size, amount, or degree of (something) by using an instrument or device marked in standard units or by comparing it with an object of known size. “the amount of water collected is measured in pints” some synonyms: count, calculate, compute, quantify

So, for all measurements offered to us as information especially if accompanied by a claimed significance – when we are told that this measurement/number means this-or-that — we have the same essential question: What exactly are they really counting?

Naturally, there is a corollary question: Is the thing they counted really a measure of the thing being reported?

For example:   Does the drastic reduction in the number of early-morning home-delivery milkmen over the last forty years really mean Americans are drinking proportionally less milk per capita? (For extra credit: The answer is YES and NO. Americans consume about 20% less fluid milk and cream than in 1975, but per capita consumption of all dairy products, including butter, cheese, yogurt, cottage cheese and others, has increased by 12.5%. For more information than you ever wanted to know, see here.)


Hungry Children?

I have seen the following statement on billboards in our area and in the press:   “1 in 5 U.S. children at risk of hunger” . Gee, one might think, that’s terrible in a country as rich as ours – and you’d be right. But the devil is in the details. It takes quite a bit of searching around on the ‘Net to find out who said that, and what it is exactly that they found that gets translated into that headline. The original USDA report is summarized here.

In fact, the original report caused a lot of push back and push-back on the push-back. The push-back link gives an idea of what is being counted here. It would not be what you think.

They did not interview classrooms full of kids to see if they “were at risk of hunger”. They did not count kids that they deemed “at risk of hunger”.

You see, there are kids that sometimes are not sure that there is going to be enough food in the home to make it possible for them to eat whatever and however much they (or their parents) want. In a nutshell, one adult in certain poor families (43,253 households) were questioned about food security, with a set questionnaire. Any family whose adult reports that they were ever worried that they would run out of money for food before the end of the month at any time during the last 12 months is counted as “Food Insecure”.

Questions like these were used to determine Food Security (each followed by the ANSWER that triggers a label of Food Insecure Household):

“In order to buy just enough food to meet the needs of your household, would you need to spend more than you do now, or could you spend less?” MORE

“In the last 12 months did you ever run short of money and try to make your food or your food money go further?” YES

“The food that we bought just didn’t last, and we didn’t have money to get more.” Was that OFTEN, SOMETIMES or NEVER true for your household in the last 12 months?” OFTEN or SOMETIMES

“We relied on only a few kinds of low cost food to feed the children because we were running out of money to buy food. Was that OFTEN, SOMETIMES or NEVER true for your household in the last 12 months?” OFTEN or SOMETIMES

In extreme cases, there are actually some children that actually missed one meal, sometime in the last 12 months, because there wasn’t enough money to buy food. This last condition — ever in the last year was forced to skip a single meal, either adult or child in the household — triggers the federal governments labeling of the family as having “very low food security”.

The true root causes of the problem, in the vast majority of cases, are single-parenthood, ignorance, and parental addictions – drugs, alcohol, tobacco, and gambling (State-run lotteries). A lot of parents living on the edge face the question: “Do I buy cigarettes or cereal for the kids?” or “Beer for me or milk for the kids?”   Far too many times, those questions aren’t even asked – drugs, alcohol, cigarettes and lottery tickets get purchased first, then the food is bought with whatever money is left over. Ignorance leads to the idea that Pop-Tarts make an adequate breakfast for school-aged children. But even in those homes, few American children actually go hungry as a general rule.

Author’s Aside: My wife and I did serious humanitarian relief work in the Dominican Republic for eight years recently, followed by another three hurricane-safe seasons in the various Virgin Islands. In all honesty, there are entire villages in which it would be difficult to find a single child who did not suffer “food insecurity” by the USDA definition every week – in fact, many children in the bateys (segregated Haitian immigrant slums) would be found to have been forced to skip at least one meal every day. In the west, over against the Haitian border, the public schools provide both breakfast and lunch in the lower grades. Otherwise, the children would go hungry until suppertime. Our program provided de-worming medication (otherwise you are just feeding intestinal parasites instead of the children) and specially formulated children’s poverty vitamins to supplement this program. They were not just “worried that they might have had to skip a meal in the last 12 months.” Teenage boys are often expected simply to feed themselves by hook-or-crook two meals a day and only offered the evening meal – rice and beans mostly, maybe a bit of chicken – at home.

Don’t get me wrong, when children are going hungry, then others – extended family, friends, community and government – need to step in, see that the children are fed and work to correct the problems that created the situation in that home.  I don’t mean to downplay the seriousness of the problem where it really exists.  No child should go through childhood hungry and undernourished — anywhere.

The point I wish to emphasize is that whenever we are presented with a very certain sounding number – like “1 in 5 U.S. children at risk of hunger” – it is absolutely necessary to ask the burning question: “What exactly are they really counting?”

The Question: “What exactly are they really counting?”       Answer: Families that worried they might or did have trouble keeping adequate food in the home for a well-balanced diet for all members of the household, for any reason, at any one time, even a single day, over the last 12 months.

The Second Question: “Is the thing they counted really a measure of the thing being reported?”     Answer: Not in the press reports…the USDA reports what it finds under its own definitions, however, those definitions do not mean what Advocates and the Press imply they mean. And what was counted is certainly not a measure of the thing that the charities imply it means when they use that headline on a billboard to raise money “for hungry children in America”.

As in the above case, when the press use numbers, the general rule of thumb is: They haven’t counted what you think (and probably not even exactly what they say they counted).  I go a bit further on any report offered to me by Single-Issue Fanatics and Advocates-of-All-Stripes – whatever they measured or counted; it probably does not really represent the thing they claim it represents.


Weather Fatalities?

NOAA’s National Weather Service (NWS) produced a report titled: “Summary of 2014 Weather Events, Fatalities, Injuries, and Damage Costs” (they do one each year). They send it out with a Press release and a nifty chart:


The red bars are for 2014, with 10-yr averages in pale blue and 30 year averages in yellow. (Those without a yellow bar didn’t start being compiled until 2005). News reports resulting from the press release point out that Rip Currents were the big killer for 2014. That characterization got me going on this.

For one thing,  57 deaths in a population of about 319 million people, which experienced a total of 2,596,993 deaths in 2014, is vanishingly small.   For example, compare this number to deaths from the flu (despite a spirited campaign to have all of us older folks get flu shots – I got mine!):


There were about a thousand times more flu deaths – despite modern medicine and vaccines – than Rip Current deaths.

So what’s to be investigated here? A tiny, inconsequential number (57) related to weather. It’s that last bit – related to weather – that interested me. How did they know which Rip Current deaths were weather related?

Rip Currents, in general, are not weather related. NOAA correctly says that rip currents are caused by:

“Rip currents are a result of complex interactions between waves, currents, water levels and nearshore bathymetry. These current systems such as alongshore and cross-shore (onshore/offshore) water motion. Along all coastlines, nearshore circulation cells may develop when waves break strongly in some locations and weakly in others. These weaker and stronger wave breaking patterns are most often seen on beaches with a sand bar and channel system in the nearshore zone. A rip current forms as the narrow, fast-moving section of water travels in an offshore direction. Rip currents can also result from a wave’s natural variability or when a current traveling along the shoreline encounters a structure such as a groin or jetty and is forced offshore. “

Did you see the word “weather” in there?  I didn’t.   I grew up in Los Angeles, California, and spent, during one year in my teens, some part of each of 200 days in a single calendar year on the beach – the proverbial California Surfer Boy. We knew rip currents. We didn’t have no stinkin’ ankle tethers in those days – you wiped out and your board headed for the beach. You, on the other hand, were in the water, swimming, at the mercy of the waves and currents – and boy did we have rip currents. We did not get carried out to sea and drown because we knew the trick. My current winter beach at Cape Canaveral, Florida, has this sign, which gives the trick:


Now, when the waves get bigger, there is more water thrown on the shore and rip currents, if they exist on that beach due to the topology of the bottom and nearby jetties etc., do generally get stronger, but they are not caused by the weather.

This prompted me to write to the NWS by email and ask them how they separated out the rip current deaths caused by weather and the rip current deaths that just happen on nice sunny days.

I asked:

“I am confused by the inclusion of deaths from “Rip Currents” in Weather Fatalities.

Two points result in my confusion:

  1. Certainly, there are more than ~ 50 fatalities from rip currents in the United States each year — it is the most frequent cause of ocean beach drownings.
  2. Rip currents are not weather dependent — according to NOAA’s page on the causes of rip currents [ definition above deleted ]

How is it that Rip Currents are listed as the cause of the greatest number of Weather Fatalities for 2014?”

I received this answer:

“Thanks for your interest in rip currents. You are correct in your first point. Rip current fatalities are difficult to track and are often under-reported by all agencies including the NWS. The US Lifesaving Association estimates there are 100+ rip current fatalities per year in the US and that is considered the best guess at a true number.

We are learning there are many other causes of surf zone fatalities as well but rip currents cause the most. Other causes of surf zone fatalities are rough surf, a phenomenon known as sneaker waves (an unusually large wave in a set which suddenly crashes onshore and catches people off guard), and other currents.

To your 2nd point, the NWS tracks surf zone fatalities to better our understanding of the dangers of the surf zone and improve our products, services and outreach. Hopefully, that reduces the number of fatalities.

Rip currents are indirectly weather dependent. Wind speed and direction create waves and influence near shore circulations and waves/near shore circulations are significant factors in rip current development.

The NWS lists rip current fatalities among other weather fatalities because rip currents are weather related and the NWS provides products, services and outreach on surf zone hazards. Hope this answers your questions and clears your confusion.”

Well, that still left me with a question: What exactly are they really counting?   So I asked:

“One final question, which I should have asked the first time: Does the NWS count ALL Rip Current deaths reported as Weather Fatalities? or only those that occur during official weather alerts?”

The answer:

“The NWS counts any surf zone fatality occurring in its area of forecast responsibility as a weather fatality.”

What exactly are they really counting?   They are counting any and all surf zone fatalities that occurs in any state, PR, Guam or the USVI.

A toddler wanders into foot deep water, gets knocked down by a wave, drowns because Mom and Dad have had one too many beers and have fallen asleep in the sun = surf zone fatality. Surfer girl gets hit in the head by the board of surfer boy, loses consciousness and drowns = surf zone fatality. Sneaker wave pushes kid against the sandy bottom, where he panics and sucks water = surf zone fatality. Over-eager boy from Kansas swims out beyond the breakers, showing off, and finds he can’t swim back in, the more he swims towards the shore, the further away it gets = surf zone fatality. Only the last one is actually due to rip current.

All of them appear in the Rip Current column in the Weather Related Fatalities chart and on the Weather Related Fatalities graphic.

So, back to The Question:   What exactly are they really counting?

Answer: Any Surf Zone fatality anywhere in the 50 States, PR, Guam, or the USVI.

The Second Question: Is the thing they counted really a measure of the thing being reported?

Answer: No, they report “Weather Related Fatalities” sub-category “Rip Current Fatalities”– they have not counted rip current deaths and what deaths they did count are not necessarily weather related.

Those 57 deaths (or closer to the suspected truer number of 100) are not all deaths which were caused by Rip Currents. Even if they had been,  they would not necessarily be Weather Related Fatalities. Even if they used the true name of the thing really counted – Surf Zone Fatalities – they still wouldn’t be Weather Related Fatalities (unless you count that they occurred on days during which there was weather, of any kind). Yet, the National Weather Service publishes official reports stating that there were 57 “weather-related rip current deaths” in 2014.

I have no idea why the NWS insists on counting what it counts or reporting what it reports – there doesn’t seem to me to be any extra profit in it for them when they call them Rip Current Fatalities instead of the true characterization as Surf Zone Fatalities. Lumping all Surf Zone Fatalities under the Weather Related Fatalities umbrella cannot be justified and the reasoning for doing so, as explained to me in the above email, is just nonsensical.

Granted, in the larger picture of government statistics, this is small small potatoes.  But, the Rip Currents Affair is a fine example of why we must always ask: What exactly are they really counting? Is it really a measure of the thing reported?


More Broadly

This point has a broader application than the two cases discussed here today. Anyone who follows any science journalism knows what kind of reporting we see.

Wild claims of certainty are made in psychology research – often based on studies done on a couple of dozen university students who are being paid to participate – or recently, a study whose results were based on children’s games being claimed to tell us about the effects of family religiosity on altruism.  In the field of psychology, the thing being counted is almost never a measure of the thing being reported, except in the minds of the researchers and like-minded psychologists.

In the health sciences, we hear of the virtues or evils of every type of food or life-style choice – studies based on statistically-derived minute differences in biometric markers for things only vaguely related to the topic being discussed and yet the health press trumpets these findings as proof that Food X is either Good or Bad for us. Near utter nonsense. They don’t actually measure anyone’s health, no less measure the health of two large cohorts, some who eat Food X and some who don’t, in a double-blind study that is capable of actually determining a health effect.

Another aside:  The Health Food and Vitamin and Herbal Supplement industries feed this frenzy of invalid attribution of benefits or harms – laughing all the way to the bank with their billions of annual sales in the US alone – 81 billion for health/natural foods and 37 billion for supplements – sales of vitamins that have been found to have no positive effect whatever for the general public who buy them, sales of foods that are no better or worse than any other foods (except one is allowed to pay a lot more for the “healthy” ones) and sales of herbal supplements that often are not actually in the bottle and whose effects on the human body are for the most part unknown.  In any other endeavor, the behavior of the companies promoting and selling “health foods”, vitamin supplements, and herbal supplements would constitute criminal fraud.


In climate science?

Well, you decide…opinions vary wildly and emotions run ahead of intellect. But the questions need to be asked, and asked again.

What exactly are they really counting?

Is the thing they counted really a measure of the thing being reported?

# # # # #

Author’s Reply Policy:   I will do my best to answer questions about the main point of this essay and its implications in a broader sense. I would love to read your examples of obvious violations of the principle involved – the principle being, of course, that one should state exactly what one is counting and then actually count that and that what one counts should actually be a measure of the thing you claim it is.

If you have a question for me, please address it to me by name, so I can be sure to reply to it.

Please, this is not about the Climate Wars (though there is a lot of this type of thing going around in the climate field – e.g., see here). While you may state your climate opinions here, I will not respond nor will I defend any particular Climate Wars viewpoint.

I hope, most of all, that this essay has made you think about the numbers you read and hear in the news in a slightly different way and to begin asking yourself these two important questions.

# # # # #


159 thoughts on “What Are They Really Counting?

    • Reply to Nicholas Schroeder ==> Exactly right — and much to the point. Whenever we are presented with a number — such as historic surface air temperature — that is really a guesstimate made from proxies, we must ask Question # 2: Is the thing they actually counted really a measure of the thing claimed?
      For example, there are a lot of solid papers that cast doubt on tree rings as temperature proxies. The thing they really measured were widths of tree rings in cores of old trees. Is that really a measure of surface air temperature?

      • It’s much, much worse in this case. NOAA is supposedly giving us numbers on weather related fatalities, but I think they counted lost emails instead. Just look at deaths related to cold, they claim a measly 29/year as a ten year average; that seems just a tad too low.
        A quick check with the CDC shows that the number of U.S. citizens that die from exposure to excessive natural cold each year averages well over 1,000.
        The CDC also concludes that three times more people die from natural cold exposure than heat exposure, yet somehow NOAA reports four times as many citizens die from heat related deaths!
        This is pure unadulterated propaganda! US government climate science* is anything but science, they don’t seek truth, they seek fulfillment of a politico-ideological agenda.

      • RWTurner December 5, 2015 at 10:01 pm

        A quick check with the CDC shows that the number of U.S. citizens that die from exposure to excessive natural cold each year averages well over 1,000.

        Good point , thanks for the numbers. You will also note that they manage to divide deaths due to “winter” and those due to “cold”. So I suppose the surprisingly small figure must be those whose who died of cold in the summer !!!
        Lies, damned lies, and statistics.

      • This entire post is about one very small thing and that is, to implant the idea that emotions: “run overhead of intellect”. But that is not the truth. Propaganda runs ahead of all other considerations today. Payed for hearsay is the dominate feature of the culture of our time. How do I know this? It is because I spent most of my working life in Communication Design. The roles of Journalist and Politician are not adversarial they are cohorts. “What Are They Really Counting?” is a very good question given you can count at all. The imaginary number of Zero is the revolutionary beginning of intelligent counting. But what would I know, I’m not Kip Hansen. Anybody that could call Willis Willis Eschenbach a troll, on this forum, and get away with it, is way above my pay grade! ;-(

      • Reply to RWTurner ==> Yes yes yes yes…..one must check: “What exactly are they really counting?”
        In both CDC examples, they are exactly counting:

        Deaths attributed to exposure to excessive natural cold (X31) (underlying or contributing cause of death or both), to hypothermia (T68) (contributing cause of death), or to both, according to the International Classification of Diseases, 10th Revision
        Deaths attributed to exposure to excessive natural heat (X30) (underlying or contributing cause of death or both), to heat stroke or sunstroke (T67) (contributing cause of death), or to both, according to the International Classification of Diseases, 10th Revision

        International Classification of Diseases, 10th Revision, usually written ICD10, comprises the codes that coroners use to mark death certificates, codes that eventually end up in CDC databases, from which these stats are derived.
        As in the Rip Currents example, Cold and Heat related deaths do not mean what you would think.
        What we want to see for Weather Related/Climate Related deaths from cold are ONLY those coded X31 “Deaths attributed to exposure to excessive natural cold”. They do not give that here. They add in T68 — hypothermia deaths. Hypothermia can occur even at mildly cool temperatures — it is the result of a body cooling off too rapidly and too far. See Mayo Clinic examples. Hypothermia can occur at any temperature below 98 degrees F.
        Similarly, heat deaths include not just those caused by excessive natural heat (X30) but also heat stroke and sunstroke (T67) which are NOT necessarily weather related.
        The point is, that while they are calling their numbers with the same or similar names — they are not counting the same things.

      • Kip Hansen says:

        Yes yes yes yes…..one must check: “What exactly are they really counting?”

        Me thinks people really ought to check what exactly are they really counting …. when they count “the number of cigarette smoke/smoking related deaths in the US each year”.

      • How exactly did they filter out the 1,000+ “natural excessive cold” related deaths supposedly not related to weather is what I’d like to know. Are they saying, for instance, if someone dies from hypothermia after their vehicle gets stranded in the middle on the Rockies in January, is not weather related? Like if the same thing occurred in July then they would have suffered the same fate?
        I’m no M.D. but I would wager a hefty sum of money that most natural excessive cold related deaths occur because of cold weather, they are saying it’s closer to 0.03%. Amazing conclusion they have come to! With that low of a correlation between a region’s climate and cold related deaths you might think that cold related deaths might be just as likely to occur in Hawaii than in Alaska. The pesky cause of death counters at the CDC again seem to think cold weather and natural excessive cold related deaths go hand in hand.

    • “Is the thing they counted really a measure of the thing being reported?”
      Ummm, no, especially when it comes to proxies of any kind.

      The length of a column of mercury, or alcohol is generally accepted as a good proxy for temperature. Obviously, you disagree. Why?

      • Rely to Pompous Git ==> (You’re not really, are you?)
        The topic of my comment way above is “especially when it comes to proxies of any kind.”
        Your question “The length of a column of mercury, or alcohol is generally accepted as a good proxy for temperature. Obviously, you disagree. Why?” does not represent the case of a proxy, but instead asks “It a length of a column of mercury, or alcohol, in a narrow glass tube a proper ‘ruler’ for measuring temperature?” The thermometer is a measuring device, and while there are valid arguments for and against various temperature measuring devices, I think it is wise to separate out that idea from what is usually referred to as a proxy for temperature.
        Using the width of tree rings in the wood of trees that are hundreds of years old to attempt to determine that the average temperature was in the area in which the tree was growing is using something not a temperature measuring device to determine past temperature — a temperature proxy. Similarly various types of measurements of the different layers in ice cores.

        • Perhaps a good informal distinction between ruler and proxy is time. A ruler measures something at a particular current moment, a proxy is a preserved condition that is considered to measure something from history, to hoped-for but varying degrees of success.
          Rulers generally focus on a single metric. Proxies generally are reflective of overall conditions and many metrics, and impacted by multiple factors that would have engaged separate rules had we been there at the time. The tree ring example is clear enough, as it involves a variety of environmental characteristics from average temps during a season to extremes of temperature to overall pattern of temperature changes (with different reactions in different environments), not to mention precipitation, humidity, altitude, carbon dioxide, soil changes et cetera. Of these, altitude does not change as much — but even so, can be enough to be an issue over long timespans.
          And we use different rulers on our proxies. What do you like, the width of the rings, the density of them, or some other aspect of their molecular makeup? Each has different implications, and it is difficult to tease good quality information about the original conditions out of these.
          ===|==============/ Keith DeHavelle

      • I must disagree with Keith DeHavelle’s choice of time as the discriminator between a ruler and a proxy. Some proxies are instantaneous (or nearly slow), others are lagged but time alone is no measure of their reliability.
        The proper discriminator is the degree of independence observed between the proxy measurement and the ideal variable under study. The length of a column of mercury is a very good proxy for temperature ONLY because we have a high degree of confidence that we have held all other variables constant and that there is very little independence between the proxy and the ideal variable.
        The length of a column of mercury would NOT be a good proxy for temperature if, for example,
        a) the partial pressure of mercury were high,
        b) the cross-sectional area of the column were allowed to vary,
        c) the heat conduit between the environment being measured and the column of mercury were subject to interference, etc.
        We trust a column of mercury as a proxy of temperature only because all those issues have been carefully studied and controls established. We choose mercury BECAUSE it has such a low partial pressure. We take great pains to control the uniformity of the column. We design the thermometer to be an efficient conductor – or at least, to ensure that any error will be uniform. Finally, we calibrate the thermometer against known reference points and (for scientific thermometers if not always for the one in your medicine cabinet) we periodically re-calibrate them to ensure that the proxy is consistently reporting the measured environment.
        So Nicholas’ opening statement that we should never trust proxies is an overstatement but only a slight one. We should always be suspicious of proxies until and unless they have been extensively tested and validated and even then we must be careful to regularly re-validate them.

      • Reply to Mike Rossander ==> Your point is well taken — assuming that every ruler is a proxy of a sort for the thing we are measuring — every measuring device thus a proxy of sorts.
        I believe you are right about the validity of a proxy — how close it comes to the actuality.
        Historical proxies, as used in climate science, are very controversial — usually with no way whatever to determine how closely they match the past actuality.
        In medical research, biomarkers, such as “blood cholesterol levels”, “red blood cell count”, BMI, the blood pressure measurements, and the like are used as proxies for “health” both good and bad — researchers have opinions on what levels are bad and what good, and any movement towards the bad side is considered to be something that “damages health” — without ever actually measuring health itself. The proxies are not, in the sense of health, valid. Patients experience no signs or sense of ill health, they live no longer or shorter, no change in quality of life.
        Counting and measurement — on a scientific level — are very complex issues.

      • The Pompous Git:
        You ask

        The length of a column of mercury, or alcohol is generally accepted as a good proxy for temperature. Obviously, you disagree. Why?

        Your question goes to the meaning of ‘measurement’.
        All measurement is comparison with something else.
        In times past the comparison was often made with variables. For example, the length known as a ‘yard’ was compared to the length of the arm of the person making the measurement and, therefore, whether a length was more or less than a yard depended on who made the measurement.
        Comparisons with variables were problematic when a measurement was important; e.g. was sufficient length of cloth for making a coat being offered for sale, or what weight of grain was in a sack. Hence, ‘standard weights and measures’ were created and all measurements were compared – directly or indirectly – to these calibration standards.
        All measurement devices (e.g. tape measures) are calibrated against the pertinent calibration standards. But the accuracy and/or precision of a measurement device may change (e.g. because part(s) of a tape measure may stretch or shrink).
        Hence, regular recalibration of measurement devices is needed if the accuracy and precision of their measurements is to be known.
        A thermometer is a measurement device for temperature. It uses the differential expansion between a tube and the volume of e.g. mercury or alcohol in the tube as an indicator of temperature. This differential thermal expansion of a thermometer can be calibrated against a calibration standard for temperature (as the marks on a tape measure can be calibrated against a calibration standard for length).
        All true measurement devices of a parameter provide indications with calibrated – so known – accuracy and precision.
        The length of a column of mercury, or alcohol, in a calibrated thermometer provides an indication of temperature with calibrated precision and accuracy.
        Proxies are measurement devices that provide indications that are not calibrated because they were not compared to calibration standards for the times of their indications; i.e. proxies provide indications with no known accuracy and precision.
        I hope that helps.

      • Been out of school for a long, long time, but I seem to remember that rulers, thermometers, etc. took direct measurements using a calibrated instrument, while proxies were an indirect measurements using calibrated instruments. Thus, a thermometer was a direct measurement of temperature using a calibrated instrument, while measuring a tree ring with a micrometer was an indirect measurement of temperature using a calibrated instrument, and was, therefore, a proxy.
        That might be too difficult for some to comprehend.

      • Jtom, that’s not quite right either. “Temperature” is a measure of average thermal energy of the system. We have no way to directly observe average thermal energy except to put our hand on it and say it’s hot. Mercury thermometers exploit the relationship between average thermal energy and the coefficient of thermal expansion of the material. By visually measuring the linear amount of expansion in millimeters (and holding all else constant), we infer the average thermal energy. So by your own definition, the mercury thermometer is an indirect measurement and therefore a proxy (albeit, a quite well calibrated one).
        Measurement is seems difficult because doing it right actually is quite difficult.

        • Thermometers as rulers (as opposed to proxies) benefit from mechanisms and circumstances that mean that the attribute desired to be measured (for our purposes, air temperature) has a near-exclusive impact on the proxy. Their could be other confounding aspects — direct sunlight, reflected infrared, evaporation, air movement et cetera — but weather stations allow for these and the result is quite good.
          This is why we have to adjust the numbers so much after the readings are taken. ];-)
          But proxies are used because no one was there to take the measurement at the time. And there are many confounding factors in things like oxygen or beryllium or carbon isotopes, depositions of biological matter, and so on. So many factors contribute that we cannot control for them, so we logic our way to the best guess … which is sometimes, sadly, the most politically useful guess.
          ===|==============/ Keith DeHavelle

      • Keith DeHavelle says:

        But proxies are used because no one was there to take the measurement at the time.

        You are right of course ….. but with one (1) exception that involves atmospheric CO2 proxies with one (1) of said proxies being extremely more accurate than all the other CO2 proxies combined.
        And those highly accurate ones are the “plant stomata proxies” because those plants were actually there at the time those per se …… atmospheric CO2 ppm measurements were being “recorded”.

    • Reply to JohnWho ==> There is a lot of good science being done in many fields. Careful experimental design and planning, metrics defined ahead of time, methods carefully crafted based on best practices in the field of endeavor…..
      So, yes, there is some scientific counting/measuring out there that is not only accurate but is plausible scientifically to be a good measure of the thing/effect under study. It is not all bad.
      In clinical medicine, there is a lot of care being taken in double-blind placebo controlled experiments that count the effect being looked at — quantifiable improvements in real conditions, full recovery of patients under treatment, large improvements in survival.

      • My bad – I was attempting to be both cynical and sarcastic regarding what is being reported by the Main Stream Media, but left out that important detail.
        Possibly science would still be in the dark ages if what you describe in the essay was the norm.

  1. There is a related phenomenon regarding numbers being reported: What happens when you plan a clinical trial including what measurements you will report, but when you’re done skip the plan and report different measurements that happen to look better?
    This sort of thing is disturbingly common, and the Compare Project has been set up to track it, call the journal on it, and ultimately publish the result.
    Rather like many climate studies (not addressed by this project, as they don’t set out a priori measurement plans), if you have the freedom to keep looking for things to measure until you find one with useful results, you’re guaranteed to find something.
    Depending upon what you mean by “useful.”
    ===|==============/ Keith DeHavelle

    • Reply to Sweet Old Bob ==> Exactly, please!
      However, the 10-year average for all weather deaths combined, however and whatever they are counting, only comes to 600 a year — and that’s for all 50 states, PR, Guam, and the Virgin Islands combined. Compare with: “In 2014, there were 726 cyclists [bicyclists] and 4,884 pedestrians killed in motor vehicle crashes.”
      The weather isn’t making much of a dent in the population.

      • 600 per year is probably less than one third of the weather related deaths that actually do occur each year.
        I’m guessing they applied the special “extreme weather” filter to get these numbers. Remember, heat is always considered extreme so applying this very “special” filter will always skew heat related deaths upwards.

  2. Reply to Keith DeHavelle ==> Very important issue. Thanks for the link to the Compare Project.
    Many fields do not, but should, require the kind of pre-registration of scientific experiments in the same way as clinical trials are registered and tracked.
    In many fields we find that even hypotheses are changed after the fact of the study — so that what should have been a prior hypothesis becomes a post hoc hypothesis, and the study miraculously confirms it!
    Much of the work in climate science, I am afraid, is done by repeated tuning of calculation parameters, without any prior justification, until the “correct” or “desired” answer is achieved. The justification for the methods employed are then added post hoc.

    • Kip many thanks. You have written the article that I always wanted to write but never did. The media drive me mad with their constant, unresearched quotes of 100s here, 1000 there. Just crude unscientific rubbish

  3. excellent reflections on counting and what’s being counted. A clear notion of those two things is something that leads to a clear and testable hypothesis.

    • Kip — Thank you for answering a question that has occurred to me before — government tells us 1 in 5 children is starving, but childhood obesity is rampant. My conclusion: many of the same children are being counted for both “factoids”.
      Last year I wrote a letter to the editor of the Proceedings of the National Academy of Science. Around the opening of hurricane season 2014, they published a paper whose authors maintained that hurricanes with feminine names are more destructive than those with masculine names. The underlying cause, they said, could only be sexism: people tend to make better hurricane preparations if the storm name “sounds” more threatening (an effect I have not observed after nearly 35 years in Louisiana).
      The paper was Junk Science of the highest order. For one thing, hurricanes were classified as having more or less strongly gendered names by a poll of a handful (11?) of the authors’ academic associates. Consequently, TS Sandy’s results significantly skewed the results because the name was judged to be relatively feminine. Sandy?
      Further, we normally associate “hurricane deaths” to mean those in areas subject to winds, tidal surges etc. in coastal areas near landfall. In fact, several of the most notorious “killer” storms in the era of named storms (since the early 1950’s) killed hundreds of people well inland (Pennsylvania, Virginia) due to flash flooding. You expect people in coastal areas to make hurricane preparations, but not in the mountains.
      I had other objections, not the least of which was that for most of the study period, storms were given only feminine names!

  4. Thanks Kip, I think we all can find some relevant memory of reading or hearing something and it triggers the BS button. The rip current discussion is a great example. Kind of like someone getting into a car accident due to a Tornado siren that scared them. Weather related?
    A rarely well understood item around “counting”, that I still find interesting, is this>
    Next up, Taxes? 🙂

    • They pretty much spell it out, but people use their statistics to try to convince us the economy is rosy when it isn’t.

  5. Kip.
    As far as rip currents, would not wind also be along the same lines. Winds and currents both act pretty much the same ways. Ocean currents, like the Gulf Stream, are comparable to the Jet Stream, though both are acting in different mediums. Rip currents are small scale events just like wind gusts from rising heat…Localized, but i would say ‘weather’.

    • Reply to Dahlquist ==> For Rip Currents, wind is only factor in that it can increase the size and power of waves. When the topology of the beach is such that rip currents are present, more waves bring more water up on the beach, which then tends to flow out in one spot — rip current. To be clear, rip currents are caused by beach topology, under the water, not by the weather above it.
      Rip currents can and do exist even on windless days and on days when the wind is contrary to the waves (blowing off-shore). It is these sunny calm days that make rip currents dangerous for the unaware, as the sea does not look dangerous. Normal people do not go in the water in rough weather (when there is really weather), mostly only surfers, sail and kite-boarders — who ought to know better.
      Disclosure: I surfed storm surf off of Huntington Beach, 15-20 foot waves — a feat of near-suicidal derring-do — but only once.
      Ocean currents are not caused by daily weather.

      • I could almost see those waves hitting Huntington Beach from where I sit, with perhaps a hundred feet more altitude and a window in the right direction. Ah, but my electric chair is not equipped for the surf. So I use other methods to stay … current.
        Your essay was quite good. Thank you.
        ===|==============/ Keith DeHavelle

  6. Good essay. I have learned simply to disbelieve when statistics are chanted in my face. I’ve seen the “One in five children…” on a billboard and didn’t believe it, but it’s good to see what’s behind it. Another point that must be made is “what statistics aren’t they showing you?” For instance, there were three clinical trials of Prozac. Two of them showed it was no better than placebo. The third showed it was marginally better. Only the third was shown to the FDA to get approval for the drug.
    Another point concerns graphs. By tweaking the scales, omitting error bars, and putting disrelated statistics on the same graph, one can suggest all kinds of nonsense. And saying two statistics are linked does not mean one causes the other.
    I can’t resist mentioning one thing in the NWS statistics: 20 fatalities from heat in 2014, way down from the 10 year average of 124, but 43 from cold, up from the average of 29. I know it’s a small sample, but if it was reversed, the mainstream press would have been all over it.

  7. I heard the statistic about “1 in 5 U.S. children at risk of hunger” in radio commercials that played all summer long. They had several children say, “I wish I was in school” to imply that 1 in 5 children would rather be in school with free school lunch than be starving at home. I don’t think you could find very many U.S. children who would actually say that during summer vacation, certainly not 1 in 5. But honest or not, the statistic served their purpose, so they used it.
    Now the ad I hear the most on day-time radio is about young men registering for the draft. The ads play several times a day, every day. I really don’t think many 18-year-olds are listening to AM radio during the day, so why are they playing the ads repeatedly? Are they planning to bring back the draft soon? If not, what purpose does it serve to spend so much government money on these ads? Are they sharing the data they collect? Is it being used for other purposes than what was intended? The media won’t investigate because they make money off the ads. But when something doesn’t make sense, it peaks my curiosity.

  8. A friend tells this story: He taught an introductory level college class and the students were to rate the class and the teacher at the end. On the general question about the teacher (using a scale from very poor, .., fair,.. to very good), one student told him something like: “You treat all of us in the same fair manner, so I marked that last question as “fair.” Still, in it went to the machine-reader and the calculations.

  9. Ronald (and Kip):
    Last May “The Lancet” published a comprehensive study of mortality. They evaluated nearly 75 million deaths over 27 years, from nearly 400 locations within 13 countries. Obviously a sufficient universe for statistical evaluation.
    One of the causes of mortality evaluated was extreme temperature, both heat and cold.. Their study revealed that about 17 times more persons died from extreme cold than from extreme heat.
    Maybe the NWS should read it too.

    • The thing that surprised me in that study was the result for Australia. I expected that cold weather mortality rates would be higher for Europe and North America having been born and raised in Northern England and lived in Ohio for a few years. That Australia exhibited higher mortality rates from cold weather AND lower mortality rates from hot weather when compared with Canada was counterintuitive.

      • All you need is sufficient water to survive heat.
        With cold, you need a source of warmth, shelter, proper clothing, etc.
        There is no place on earth where an adequately hydrated person will quickly die of heat.
        Large portions of the planet, including the majority of the ocean surface are, for some or all of the year, cold enough to quickly kill an unprotected person.

      • Reply to Menicholas ==> Consider that thousands of people intentionally enter steam rooms and saunas to sit or lie for an hour or two in temperatures, and humidities, far above anything Nature throws at us. I used to do it nearly every day, and when traveling for IBM, used hotel saunas every night.

      • The finding (about differential mortality between Australia and Canada) is not surprising to me at all. And I again think that it is a problem with the connection between what’s measured and what’s claimed to be measured, though the problem is perhaps a bit more subtle in this case. The problem is their definition of the “weather event” that qualifies as a trigger to start counting. The definition is generally a set number of degrees above or below normal when what they should be using is standard deviations away from normal for that area. Australia is generally hot and Australians are well prepared for more of the same. Canadians (and us north-east Ohioans) are used to cold and ready with the clothing, backup heating systems and snow plow services to survive more of the same. Give us an unseasonably hot summer, however, and you discover the hard way that your air conditioner is substandard (if you have one at all) and that you don’t have the cultural and dietary traditions that help populations in warmer climes avoid heat injury.

  10. Being selective about the data used to justify an argument is endemic. The pursuit of personal agendas using skewed or irrelevant “facts” is commonplace – politicians, marketing companies, climate sceptics and activists (to name but a few) are all at it.
    Charities are in many ways the worst offenders – using duplicitous and questionable statistics can easily undermine any faith the public may have in their good intentions. If one were to add up all claims of populations suffering from serious chronic problems there would likely be no one healthy left!!
    We should also adopt a healthy scepticism when presented with these types of argument, make common sense assessments of their reasonableness, and seek alternative corroboration. Sadly my first response when presented with such arguments is to doubt their veracity, rather than be convinced by the case made.

  11. BBC has a radio show/podcast on this very topic, understanding if numbers used in news etc. are what they seem:
    More or Less : Tim Harford explains – and sometimes debunks – the numbers and statistics used in political debate, the news and everyday life

    • Reply to climatebeagle ==> Thanks for the link — I don’t get BBC radio here in Florida or in Upstate NY, nor in the islands, didn’t realize I could listen online.

  12. I give to the local food bank, as they provide supplemental meals to the homeless and indigent, who, I don’t care what your policy thoughts are, shouldn’t starve. That said, I live in an old neighborhood which today can charitably be called “working class”, though I have doubts on the “working” part, in many cases. All the children are, sadly, obese, and nearly all the parents as well. My neighbor died recently of complications as a result of type 2 diabetes, which could have been prevented with simple weight loss. Lack of food in the US is not a problem. The commercials saying 1 in 5 children go hungry, are, on their face, disingenuous. The same misinformation and twisting of data are now endemic in all reporting to achieve the desired spin. Any subject, any phenomenon, is twisted to agenda. I don’t know how one fixes this, but the authors point is spot on. Kudos for a well written essay on the subject.

    • Reminds me of the conversation between the Enron exec and his accountant:
      Question: How much is two and two?
      Answer: What do you want it to be?

  13. [W]hen the press use numbers, the general rule of thumb is: They haven’t counted what you think (and probably not even exactly what they say they counted)

    Amen. Although probably almost all of us who have reached a certain age have also reached that conclusion, I’m not sure we always keep it in mind, so it’s good from time to time to be reminded.
    As to the one-in-five-children statistic, my doctor brothers tell me that, outside of drug abusers or anorexics, they have never, ever seen any malnourished patients in the U.S.A..

    • Joe, I was typing when your comment posted, but it appears your brothers and my GP are in total agreement.

    • Joe: perhaps malnourished kids don’t see a doctor because the parents can’t afford health insurance?
      This is how many of us in other countries imagine life in the USA, by the way.
      That should provoke some serious wrath from the right-wing contingent!!!

      • Reply to Smart Rock, Windsong, and Joe ==> There are malnourished children in the United States but they are victims not of a general state of poverty, but of criminal child abuse — usually abused by neglect by drug/alcohol addicted parent(s). When these kids get to school age, they are identified and most often removed from the abusive parents — yoyo’d in and out of the parental home as the parent undergoes attempts to sort things out and end the abuse. Almost none of this has to do with simple poverty. As I mentioned in the essay, the real problems are almost always single-parenthood, ignorance, and the parent’s addictions.
        These are the kids that child welfare agencies focus on — not the USDA’s “at risk of hunger” kids. There is value in the programs that help lower income parents afford enough — and the right — food for their kids — WIC, SNAP, etc. Most religious organizations — churches — have programs to help in their communities. All of these serve a worthy purpose. The best of them include addition treatment for parents, education on nutrition, personal finances/budgeting, and job training and job placement.

        • @Kip Hansen, who wrote:

          There are malnourished children in the United States but they are victims not of a general state of poverty, but of criminal child abuse — usually abused by neglect by drug/alcohol addicted parent(s).

          Indeed there are. One of these was trotted around the country by Hillary Clinton to demonstrate the need for the Clinton health care plan (“HillaryCare”) in the 1990s.
          This scam fell apart when the girl was discovered not to be a sickly and malnourished patient on poor healthcare, but was instead the victim of criminal child abuse. Hillary’s partner in this scam, the girl’s mother, did years of jail time as a result.
          The girl’s mother is out now, but Hillary was never charged.
          ===|==============/ Keith DeHavelle

      • I started life in a wealthy European country living in government housing. We had 4 of us sleeping in a small room that doubled as the common room and kitchen. 2 tiny bedrooms contained 2 more people apiece. My grandfather was the long serving town drunk.
        4 of us immigrated to Texas where we lived in a new house trailer that looked like a palace to our eyes. Returning to Europe over the years I have never had cause to regret the move. So yes, when I hear the home of my choice slandered I instinctively rise to it’s defense.
        Ignorance is what we try to avoid here. I think it’s what brings us together.

      • there is value in the programs that help lower income parents afford enough — and the right — food for their kids — WIC, SNAP, etc.

        Perhaps. But it seems to me that the low-income person’s problem is lack of money, whereas that of malnourished people is not eating right. Two different things. If they don’t have enough money, give them money if you want and let them decide how to spend it in such a way as to maximize their utility. If they aren’t eating right, make instruction available about how to do so. If parents aren’t taking care of their kids, have child protective services intervene. Have the solution fit the problem.
        But alphabet-soup programs and associated bureaucracies seem to me hideously inefficient ways to go about things. It would be better to eliminate all means testing, all income-tax-rate differences, all tax credits, all voucher programs, all school-lunch programs, etc., in favor of one department of the dole: one-stop shopping for welfare money. You fill out as many pages of the application as you want, submit it to the department of the dole, and get a check. Then If your children aren’t eating right, you get investigated, because you probably could be providing for them but aren’t.
        As a bonus we’d know as a society how much we’re spending on welfare.
        Okay, I got that off my chest. I feel better now.

      • 1. You would expect death to be one potential outcome of malnourishment, and the death of a child would be cause for an autopsy in the US. You won’t find malnourishment as a contributing factor in very many, if any, childhood deaths. If you do, you will also find ‘criminal negligence’ along with it.
        2. Most everyone knows that there are free school lunch programs throughout the country for those in need. What is lesser well-known is that there are free school breakfast programs as well.
        3. Because of #2, government provided food vouchers (food stamps, EBT cards WIC assistance, whatever) really only need to provide for one meal a day, supper.
        If there are malnourished kids in the US, it is because they are selecting what foods to eat with little or no adult supervision, and filling up on French fries, chips, soda, and candy. It is not from a lack of proper food available to them.

        • On child hunger in the US: As I write research and other grants professionally, I get lots of emails from philanthropy-oriented sources.
          Here’s a bit of news received moments ago: “Hunger is Still a Huge Problem in the US. Who is Funding to Promote Food Security?
          The source, Inside Philanthropy, notes these statistics:

          While many families are buying all the extra fixings to make Thanksgiving dinner special, 79 percent of low-income households in Feeding America’s client base report “purchasing the cheapest food available, even if they knew it wasn’t the healthiest option, in an effort to provide enough food for their household.” We also know from Feeding America’s report, Hunger in America 2014, that food insecurity has been on the rise since the Great Recession: One in seven Americans rely on food banks to see them through. Viewed by race, the results are even more startling: One in four African Americans relies on a food bank; one in six Latinos. Meanwhile, some 45 million Americans rely on food stamps.
          It’s 2015, and hunger is still a huge problem in America. And it’s a problem inextricably linked to larger issues of economic hardship. In fact, many working Americans face food insecurity, with studies finding that a growing share of food stamp recipients participate in the labor force. This is part of a broader story of the difficulties that low-wage workers confront in making ends meet. Earlier this year, a study found that about 48 percent of home healthcare workers are on public assistance, as are 46 percent of child care workers and 52 percent of fast-food workers. Another big category of hungry people are older and disabled Americans on fixed incomes that fall short every month.

          It is just barely possible that there are other explanations for these numbers that would better describe the situation instead of “food insecurity has been on the rise since the Great Recession.”
          An example: As I am an older person, the local community government is pushing to bring me meals every day. They don’t care what my income is, and will charge me $7.50 a day (for three meals) if I can afford to pay it. But this is voluntary. No doubt, I would end up on such a supporting statistic if I gave in to the advertising.
          These days, being “on the dole” is not shameful, it is fashionable. And easy. And quite adequate for the needs of many, who then are retrained out of whatever work ethic they might have once had. But they are measured, yes indeed, and people accepting generous handouts are a common (if ill-advised) proxy for “food insecurity.”
          ===|==============/ Keith DeHavelle

    • Perhaps “mal-nourished” is the wrong word in the context and “under-nourished” a better alternative. The pre-fix mal- means badly. Obesity is a result (often) of mal-nourishment not through lack of food but of unsuitable diet so that doctors claiming never to have seen mal-nourished patients (outside drug-abusers and anorexics) perhaps need educating in the meaning of words.

      • As it happens, I deal with chronic malnourishment personally. But this is from having lost a portion of my small intestine decades ago. Food doesn’t get absorbed very well, especially fats. On the good side, I can eat (and enjoy) a lot of meals and not gain weight. And my dozen eggs a day keeps the cholesterol down. I just have to take various supplements to keep blood numbers good.
        ===|==============/ Keith DeHavelle

  14. This comment is in regard to the “Hungry Children?” section. About five years ago, I was in an exam room of the large local medical system waiting on my GP. Was thumbing through the monthly magazine published by that system. On the cover was a photo of Jeff Bridges accompanying a headline for an article about eliminating hungry children in America. (Around 2010, Jeff Bridges became a spokesman for the No Child Hungry campaign.) When my GP arrived, I pointed to the cover story and asked if he had ever seen starving, or even hungry, children in his practice. He takes a quick glance at the cover and replies, “Never. But I sure see plenty with the opposite problem.”

    • My two sisters are both nurses in the UK and while they have seen malnourished children very few are from families too poor to buy food.
      The largest group are in fact from middle class well educated families who’s parents have decided to become Vegans. While a vegan diet is perfectly sustainable for adults it really easy to end up with children who have serious health issues due to eating foods that simply don’t provide enough protein, omega 3 fatty acids or vitamin B12. The latter was found to be the cause of marked rise in the incidence of pernicious anaemia amongst children.
      The second largest group is among poorer people who have simply never learned to cook. The reality is that I see many people who spend more on a single meal of a big burger and fries or TV dinner than they would buying a bag full of vegetables and cheap cuts of meat that would feed them for days. You don’t even need a cooker, a simple crock pot that you can slow cook this stuff in is a wonderful investment.

      • “The second largest group is among poorer people who have simply never learned to cook. ”
        It may be even worse than that. The soup kitchen my wife volunteers at additionally offers the patrons food to take away with them. When the offered food is unsliced bread, the patrons won’t take it.
        I can only speculate at how lacking in resourcefulness a person at risk of hunger would have to be to turn down free bread because it isn’t sliced yet.

  15. Thanks for this focus on determining exactly what is being counted. Since the economy collapsed in 2008, I have kept track of government figures on such things as the numbers of unemployed and the rate of inflation. I also do my family shopping, and have seen the huge rise in food costs. It appears the counters don’t include exact foods, since if they report the true loss of purchasing power for the average person, they would have to increase social welfare payments like social security, by the percentage of inflation…..Most people who don’t shop trust that inflation is nil…Those who do shop know they have lost purchasing power……
    Also, the ‘unemployment rate’ apparently counts only people who have signed up for unemployment, not the actual number of unemployed persons who would like to have a job…. It took me a while to uncover the truth about the government stats….

  16. Some attribute this quote to Einstein.
    “Not everything that counts can be counted, and not everything that can be counted counts.”

    • Reply to Gunga Din ==> Yes, certainly true, but incorrectly attributed though — it is unlikely to have been said by Einstein — certainly never written in any of his published work and the claim that he wrote it on the blackboard (or had it printed on a sign hanging on the wall) in his office is unsupported.

  17. Many years ago, reading, IIRC, “psychology to day” or some such at the doctor’s office, I came across an article (auto exhaust and smog were topical) where a psychologist investigating California highway accident frequency was checking out his idea that concentration of carbon monoxide caused slower reflexes and irritability in drivers thereby causing accidents. He set up a CO monitoring device by a major highway and lo and behold, he was able to show in a graph a correlation between the concentration of CO in the air and the number of accidents along some tens of miles of the route! I kid you not.

    • Reply to Gary Pearse ==> Astonishing finding — traffic accidents happen mostly on roads with cars and trucks!
      Andy Revkin, host of the NY Times’ Dot Earth blog, was recently suckered by a similar claim — see here.

      • “Are there a large number of VW diesel owners in the Channel islands?”
        I was out there on Thursday and didn’t see a single one…. 😉

      • Well that’s not quite true. One of the guys I was with owns a VW, so I did see an owner out there. His car, however, was parked in a garage in Oxnard for the day.

    • Wrong way causation, Gary. Local CO is caused by traffic. What the CO monitor is measuring the amount of fuel used in the area. Not only are accidents more likely in heavier traffic, but when a wreck happens, you get a lot of idling in the area.
      We had a major problem for years in the Houston Regional Monitoring System, as poorly sited monitoring stations would give wildly inaccurate numbers for the area due to our legendarily bad traffic jams.

    • Wouldn’t be a difficult correlation.
      Higher CO due to more traffic, more traffic then a greater likelihood of an accident.
      However it is possible that greater concentrations of CO might have the effects the psychologist suggested, or lower levels of oxygen or flocks of blackbirds.
      If you don’t design your test properly its the old gigo…..garbage in, garbage out.
      Al Gore didn’t understand this either.

  18. Author’s Aside: My wife and I did serious humanitarian relief work in the Dominican Republic for eight years recently,

    Commenter’s Aside: Respects to you and your wife.

  19. NWS reported on rip tide fatalities in graph that there were 57 in 2014 and 51 average over the last 10 years. They told you in message that 100 was the true yearly number.
    So isn’t it just as likely that the reason that 2014 in above average is that the reporting was better in 2014. They think that only one in every two is reported.
    Of course, the 100 number was pull out of somewhere and I don’t think it was thin air. More likely a darker source.

  20. That is how I became a skeptic. I didn’t know much about climate but I heard a news report on the net that 70 billion tons of Ice just melted in Antarctica. It was cold out and I was a bit surprised. The news report showed a climate scientist looking at a wall sized map of Antarctica with red arrows sweeping around it signifying warm water. The reporter said that we can expect more of this with Global Warming. I decided to look into it and the first thing that I realized is that I live in the northern hemisphere. So winter to me is summer in Antarctica. One might expect ice to melt in the summertime. Then quickly I found out that indeed 70 billion tons of ice just melted in Antarctica and that was the smallest amount of ice ever recorded melting in the summertime in Antarctica. So, indeed, exactly what were they measuring and what did it signify?

  21. T-Braun: “….It’s not only a question of what is being counted, but who counted it….”
    I disagree……l believe it is why it is counted that is most important.

  22. One of my Atmospheric Physics professors once told us, and I quote “Context, Context and Context”. Be precise in context, Be clear and deliberate in context And above all, communicate that context clearly.
    Most lay people who read scientific literature are oblivious to context and asking specific relevant questions to gain context. Without context It is easily abused and twisted for agenda.
    What are we counting again and what is it we are attempting to show or define?
    Excellent essay! One that even old dogs need occasionally, to remind themselves of the old failings that we need not repeat by being lazy..

    • Reply to Bill H ==> You are playing my favorite record: “Context, Context, Context” that old favorite or of pragmatists everywhere.

  23. “unless you count that they occurred on days during which there was weather, of any kind”
    Well, yes, of course. If there is weather, the death is weather related. And a lot of weather makes up climate. So the death is climate related. And climate changes. So the death is Climate Change related. And Climate Change is, as we all know, caused by man-made CO2. So the death was caused by man-made CO2. QED.
    An unanswerable argument, isn’t it?

  24. Just ran a search on this page for “valid”, no occurrences. The issues discussed here are usually framed in terms of validity vs reliability. This article lays the validity issue out without referring to this highly technical term which most journalists and political leaders don’t get. Unfortunately many scientists(loosely used), seem to have a similar problem.

    • Reply to Chuck Susmilch ==> I have avoid the issue of “validity” intentionally — as it is a subjective context-determined judgement, in order to focus on the two questions asked here.
      The second question, of course, entails the concept of validity when it asks “Is the thing they counted really a measure of the thing being reported?” — at all, you see? If it is at least a measure of the thing, then we can go on to address validity and applicability.

      • Was trying to give you credit for discussing validity, which no one seems to be concerned with, certainly not the press or politicians or government functionaries.

      • Reply to Chuck ==> Yes, I realize that…just clarifying why you don’t see the word in the essay.
        We could measure temperature with one of those floating glass ball decorative thermometers they sell at museum stores….but using such to measure temperature to the single degree would not be valid. It does at least measure what we mean by temperature….just not valid for our purposes.

  25. 1 in 5 American children at risk of hunger- sounds incredibly low. Children have their stomached stuffed so much that they are never hungry no wonder you have high rates of obesity.

  26. I know I have seen data/graphs showing a downward trend of deaths/injuries in the last 50 – 100 years globally from extreme weather events.. Have seen them on WUWT, but can’t seem to find them…
    Would be a good graph to post here…

  27. Very well said.
    When I was young in Australia it was a mark of education or just common sense to treat statistics as inferior knowledge at best. A Test batsman’s average did not matter until you knew how and against whom it was achieved. Somehow, statistics were a bit too “yank”. (No, I’m not blaming America for our Lewandowskys, okay? Australia is now sucker-central.)
    What is accepted as academic survey now would have been push-poll then (if we’d had the expression “push poll”). It was obvious that anyone could ask a question in a way to provoke desired responses. That was for selling soap, not for serious discussion, even if you were just a bloke in the pub talking footy.
    So much science now is just speculation based on slob terms (like “global warming” or “climate change” or “food insecure” or “death in custody” or “surf zone fatality”) and then expressed in as many convincing numbers as you can pile. One of the worst stitch-ups for me would be the records of past “temps” based on min/max. It seems not to have occurred to our climate botherers that min/max is too often just a record of how high or low the temp was allowed to go by the presence or absence of cloud at a likely peak or trough time of day. I find that a staggering oversight. But do they even care? Temp records don’t need mangling – they’re born mangled.
    Anyway, Kip, thanks for your hard, persistent, frustrating giving. I know a few hard-heads around here who have no illusions about (or thanks for) their work in the local indigenous community but who keep backing up and trying, while opinionators like me just sit the sidelines. You and your missus are the real Test players.

  28. “Not much science”? There’s a lot of science in there – it’s just a different kind. Well done.

  29. In UK reporting its constantly 1 in 3 or 1 in 5 of whatever handwringing issue we need to get into. Get yer £3 a month remittance out. Its rolled out like that so often that I simply ignore. No qualification (nothing adequately validated) about the entity measured in most instances. And if a sample on anything is revealed I keep saying every time…”neither I or my wife were asked”. So who/where are these people that allegedly were polled/asked?
    Of course no follow up reported…just run the next handwringing issue out. Much of it is from charities that I am deeply suspicious of. Just say RSPCA and go from there….!

  30. Here’s a numbers story from the UK which may amuse you. I don’t know if it’s relevant to other nations.
    On our motorways ( major freeways) the inside lane – or ‘hard shoulder’ – is forbidden to drive on. It’s reserved for broken-down vehicles and for emergency service access. When you see folk broken down on the hard shoulder, you will notice that almost invariably, the occupants, young and old, fit or infirm, are OUTSIDE the vehicle, standing disconsolately around shivering in the cold and the rain, for as long as it takes the rescue truck to get there. Why? This was a puzzle to me until I heard that it was standard police advice, and indeed government advice, to leave the vehicle in such circumstances.
    And the basis of that advice? This is what the police will tell you –
    “Do you know that, ON AVERAGE, the time it takes for a vehicle parked on the hard shoulder before it gets hit by another vehicle? ..TWENTY MINUTES” . I’m not sure I have ever seen such a stunning misuse of statistics, at least not one that has translated into policy. I’ve pointed it out in a letter to the motoring press, and even they didn’t seem to get it. What hope is there for the general public when they get snowed by numbers in the climate debate?

    • The figures reported by the police and highway authorities was that the average incident that resulted in death that occurred on the hard shoulder happened within 30 minutes of the person arriving on the hard shoulder. The usual mob of illiterates in the press then garbled it.
      The advice from the police is simple
      Exit your vehicle via the side furthest away from the traffic
      Call for help via a motorway emergency telephone (mobile phones do not pinpoint your position or link directly to the motorway control).
      Never attempt to repair your vehicle on the side of the motorway
      The safest place to wait is outside your vehicle, behind the hard shoulder’s crash barrier
      Having seen a car hit by a truck while on the hard shoulder of the M1 I can attest to the fact that the advice is excellent. It is a matter of record that each year 6-8% of all fatal accidents on British motorways involve vehicles that have broken down.

      • Reply to Keith and Mothcatcher ==> Thus the value of full disclosure on “What are they really counting?” — with that knowledge, sound reliable FULL advice leads to a logical conclusion, including, ” outside your vehicle, behind the hard shoulder’s crash barrier

      • Thanks for the safety-first advice, Keith. I wasn’t really aiming to assess the overall wisdom of the police advice. Of course the hard shoulder is a rather dangerous place (although I’m not at all sure that the fatalities you cite are disproportionately amongst those who sit tight in the vehicle – most vehicles can withstand a pretty big hit, especially if there’s no object in front of them, and there is by no means always a crash barrier, and when it’s there is not of sufficient height to ensure prevention of a truck override)
        The joke was about the statistics – the only thing you can infer from the 20min (or 30 min) quoted is that it is approximately half the average time that it takes the rescue truck to get there.
        But I think unless the weather is particularly clement, I’d take my chance in the vehicle. There’s risk and risk.
        Sorry slow response, just spent much of the day on the M3/M25..

  31. Bean counters. Bear counters.
    Btw….the Polar Bears numbers are increasing. Government bear counting bean counters still insist they’re declining tho. Churchill’s bears are busy hunting on the ice, for seals and bean counters.

  32. The answer you get depends on how you phrase the question(s)…

    Unfortunately I can’t find this that funny any more.
    What occurs to me is the extreme danger of having a mass of the population, ‘informed’ by mass media, but with absolutely no ability to discriminate between competing views, especially on matters of maths or science, and with even less ability to understand the metaphysics upon which the claims and counter claims are based, even at serious doctorate student levels.
    And this article is about metaphysics, even though it never uses the term…
    ‘What exactly are they really counting?’ is a statement that reeks of metaphysics.
    1/. What to we mean by ‘Exactly?’.
    2/. Who are ‘They’ ?
    3/. What do we mean by ‘Really’ ?
    4/. What do we mean by ‘Counting’?
    It is the consensus metaphysics – what we all agree is ‘approximately true’ – that determines the nature of these words, in the context.
    The game of the Left is to invisibly change that metaphysics, by sliding in politically loaded assumptions into our consciousness, that bend the debate towards their agenda, even if it doesn’t mean they win the argument, the meta-agenda is to establish what the agenda is.
    This whole site is a tribute to the effectiveness of that meta agenda. I mean, what is the point of us wasting huge amounts of time discussing climate change, when all the evidence is that its nothing to do with us and there is nothing we can do about it?
    The sceptics will have won the war when mentioning AGW simply evokes a puzzled frown ‘wasn’t that something to do with climate that that Luddite movement, the Greens, tried to con everyone with 50 years ago’

  33. I find the Winter and Cold numbers incredibly ;ow. In the UK there is a Government report on excess winter deaths every year.

    An estimated 43,900 excess winter deaths occurred in England and Wales in 2014/15; the highest number since 1999/00, with 27% more people dying in the winter months compared with the non-winter months.

    I can only assume that either the report Kip Hansen is talking about only includes people slipping on ice and breaking their necks or falling through ice on frozen lakes and rivers and drowning and the like, and that in the USA no old person dies of winter related illness or cold related issues.

    • Reply to Sandy In Limousin ==> Here you address the crux of the issue: “What exactly are they really counting?” in the US? and in the UK? You are right that the disparity in the numbers points out they are counting different thing, even though they call them by the same (or a very similar) name: “excess winter deaths”.
      That’s why we ask the two questions.

    • Just a guess from a Canadian but winter here could involve “excess ” deaths from bad driving conditions, snow shovelling heart attacks, influenza in old people, carbon monoxide poisoning from defective heaters, etc. It regularly gets to -30 C where I live and we even get the occasional freezing death. 43,900 is a big number. I can’t believe that doesn’t include flu deaths which have very little to do with temperature and much more to do with people being inside more and in close proximity.

  34. The “counting question” I’ve always had, and haven’t had answered satisfactorily yet, is “How does anyone know that the gridded temperature extrapolations for the polar regions is even close to accurate?” I asked this question once before, and the only reference I got was to one paper that showed a correlation between measured temperatures and extrapolated temps as latitude increased — basically saying that yeah, it gets colder as you go away from the Equator.
    Yet these extrapolated temps show up with the same precision as those measured directly. So, I’ll ask the question again: Where is the evidence that those extrapolated temps for polar latitudes have any accuracy at all, given that we’re measuring Earth’s actual temps and not an ideal sphere with perfectly gradient temperatures across it?

    • Reply to James Schrumpf ==> “temperature extrapolations” are NEVER accurate, anywhere. It is a contradiction of terms. Temperatures, where not measured, are not known. The average temperature of a place where temperatures were not measured can not be known. If they are not measuring, if they are not counting, by some method, then there is no measurement to report and no count to report.
      Instead, they report their “temperature extrapolations” and use them as if they had been measured or counted. This may or may not be scientifically justified.

  35. I have no idea why the NWS insists on counting what it counts or reporting what it reports – there doesn’t seem to me to be any extra profit in it for them
    Flanders and Swann did a sketch called ‘A Transport of delight’ in which it was stated that travelling by air was much safer than by road. Flanders observed that crossing the road near where he lived was very risky what with all the airline buses whizzing around, and he suspected the drivers had instructions to ‘Keep the statistics favourable’.

  36. Deaths due to weather is a very inaccurate measure of what is stated. This gem:
    “At least 23 people have died this week in the United States due to winter weather, CNN has confirmed.
    A majority of the deaths –18 — occurred in Tennessee.
    Three of those people died in a fire in Knox County, Tennessee emergency management officials said.”

  37. One of my favorites are the medical/health care studies which measure overall mortality in a group taking x or y drug or not taking it and then claiming that taking some specific drug will extend one’s life, irrespective of the purpose of that drug and the actual cause of death. I would cite statins as an example, which seem to be vastly over prescribed.

  38. Very informative and coherent post Kip!
    I’m sure many of us can relate to a wide variety of topics in which we notice this type of “counting” is obvious, and very misleading.
    I’ve come across several in diverse topics , but when I’ve taken the time to ask for an accounting (or to point out the apparent discrepancies in the tally), my query is generally ignored or, I’m branded with an ad hom in response.
    That’s what makes a post like yours so vitally important.
    It’s sometimes humorous.To my right-leaning acquaintances I’m a “greenpeace/save the whales /hippie farmer/anti-vaxxer/truther (and worse)”, and to my left -leaning acquaintances I’m a “friend of George Bush/Big Oil/ Big Pharma/ Big Tobacco/Big Ag and the general military-industrial complex (and worse)”.
    No to both. I just don’t like nonsensical information no matter where it is coming from or what it is being applied to. 🙂
    Thanks for taking the time to write and post this!

  39. This morning a major West coast newspaper ran a story quoting Environment Canada saying the recent wet weather was unusual. The comments section lit up with people who have lived on the west coast for a long time. “This is typical fall weather. Stop the fear mongering’ was the almost universal posting. I was going to post the link but for some reason the entire story is gone from the newspaper on-line edition. hmmm

    • Reply to Jim Barker ==> Thank you for the link to John Brignell’s site == I have followed him for years.
      For many years he faithfully kept the Warm List — but has stopped adding to it (see the very bottom of the linked page for the reason).

  40. I forgot to respond to Kip’s request for examples of “dubious counting”.
    The first was my attempt to compare the effects of “second-hand alcohol (ETOH)” to “second-hand smoke” and their relative impacts on societal health and cost, for my very first statistics project some years ago.
    I learned that the “second-hand smoke” numbers are largely made up; second hand smoke effects were attributed even in populations where there were no smokers at all, via “estimates” (a lot like “surface temps” are extrapolated.). I had to give up on it, due to the sheer amount of time it took to ferret out even a few of the actual/true second- hand smoke figures (they really are not what has been promulgated, folks) , it would have taken more than six months! So, chose instead to do a project on what it actually meant to be “on time” at school (this was before most people had cell phones LOL)–answer “not much”!.
    In addition, what constituted a “smoker”, was someone who had ever smoked a cigarette , even if that person smoked only one at age 15 and was now dying of kidney failure at 90, it was given a weighting factor in that person’s demise.
    On the other hand, someone could have been drinking a quart of vodka a day since age 15, never stopping until they were admitted to the hospital dying of brain damage, cirrhosis, or heart disease at age 50, yet if the person hadn’t had a drink in three weeks, it wasn’t considered a health “factor” in that person’s demise. Crazy!!!
    Second-hand alcohol effects and cost are present in many ways throughout our society, some of the most obvious are Fetal Alcohol spectrum disorders, child death, abuse, and neglect, elder abuse and neglect, domestic violence in general, traffic injuries/fatalities, incarceration rates, psychiatric hospitalizations, murders and suicides, gun, knife, drowning /boating fatalities, etc. These are in addition to the direct physiological health effects on the person who is drinking..
    Another example is when I started looking into claims that “Mexican farmers were being driven off their land by the introduction of GMO corn there” and “big ag was driving them from their land”,etc, offered as an explanation for why so many illegal immigrants from Mexico were flooding over US borders. I kept seeing similar claims from various sources such as Oxfam, etc. Again, it took a lot of time to even come close in trying to determine where the figures and subsequent claims were coming from. (Just exchange the notion of “big ag” refugee for “climate refugee” and the picture becomes clearer ).
    The only figure I ever found that even seemed remotely reliable was that only 25% of rural Mexican citizens actually farmed in some way (often on property owned by another), the other 75% made their living providing other types of services to the rural community. These figures covered the period from approximately 2000-2010. The true biggest factors for the mass Mexican exodus that I found appeared to be, from multiple fairly reliable sources, was that Mexico, after having achieved stable population and economic growth in the 1970’s , experienced an unsustainable population explosion through the 80’s and 90’s (when the legal “age of consent” for females was 12 years old, contraception/birth control was not considered an option, and the average family had 7 children) along with the decline and essential collapse of the Mexican economy, leading to to slim-to-no economic opportunities or options for today’s younger and considerably poorer generation of Mexican citizens. And “Big Ag” or “GMO’s” had nothing to do with any of this.
    By the way , I’m usually called a “racist” when I report this information, even though sensible people know that Mexican citizens are of every color and ethnic background, even if the dominant culture there is “latino”. Further, I do think that the activities of “big ag” warrant our close scrutiny, and that paying attention to the science of all GMO’s is worth our effort. Some are clearly more beneficial than others.
    So these are just a couple controversial “counting” topics I have personally investigated-there are several more.

  41. Author’s Postscript:
    Thank you to all of you who took time out of your day to read this essay and to those who made the effort to join in the conversation or state their views in the Comments. Many good examples of the failure of some to clearly state exactly what is being counted/measured and the failure for the thing they counted really a measure of the thing being reported.
    Lots of good viewpoints and a much appreciated lack of trolls and trolling behaviors.
    The point has been raised about the validity of the final measurement or the count. For instance, they count counting bears, clearly stating that they are counting counting bears, and show they are actually counting them one by one by taking them individually out of one bin and placing them in another bin until the first bin is empty. This answers our first question: we know exactly what they are really counting and are assured by the explained method that they are really counting what they say.
    How else could they have counted? They could have weighed one counting bear. Weighed an empty bin. Placed all the counting bears in the empty bin and weighed again. Subtracted empty bin weight from total weight to arrive at weight of counting bears, then divided by the weight of their single sample bear. Their final figure is 99.5 bears. Well, OK, round to 100 bears and call it a done deal.
    Is that a valid method of counting counting bears? Not for a six year old being taught to count to one hundred, he really needs ten rows of ten each counting bears. And certainly not if they are solid 24-karat counting bears — there had better really be 100 and not 99 in the shipment.
    It might be valid if one is selling counting bears wholesale, in multiple gross lots loose in plastic bags — ten gross plastic counting bears per bag — $19.99.
    I have a couple of things to do first, but I hope soon to have an essay about the validity of measurements. Tentative title == “Apples + Oranges = Fruit Salad”.
    Thanks again for your time.

  42. Paying people in government to count hungry children darn well guarantees that there will be hungry children as much as paying people in government to study global warming darn well guarantees that there will be ….

  43. Your discussion unter the heading “More Broadly” sheds a lot of light on the climate change buisness. In our society, where people want to bleive BS, it is really no supprise that the alarmists are able to make bleivers of the masses.

  44. I first encountered the “1 in 5 children go to bed hungry in our area” claim when I saw it years ago on a Second Harvest billboard. Now Second Harvest is a nice organization which no one wants to harm, but I really doubted the claim. Because if it were really true, every minister, every professor at UT, every politician in the area would be preaching about it constantly. And I wouldn’t see the masses of fat kids every time I volunteered at my kids’ public schools. So I checked. Turns out the claim was supposedly based on a study where they asked kids if they had ever gone to be hungry.
    Ever. Heck that would have to include me whenever my mom fixed tuna casserole.
    {If they “ever” feared “going to bed hungry”. If they ever “worried about food” the previous year. .mod]

  45. Kip, there is another aspect of rulers and proxies not addressed in this post nor in any comment so far: Averaging. I’d bet you would do an excellent write up on it.
    When we look at a proxy for temperature, whether tree rings or oxygen isotope changes in diatoms or whatever, these are presented in graphs typically as fifty- or one hundred-year averages. But what is being measured, then? Do we think that no individual year is a useful measure? But we think that if we put enough of them together, it’s a valid proxy?
    I think that in many cases the proxies have some degree of validity, and that we are able to discern individual layers/rings/years adequately enough. But the result, reflecting ever-changing conditions, is noisy.
    Just like modern weather.
    It is one thing to say that 1999 was hotter than any century average in the last x thousand years. (Such statements are often untrue in the absence of additional manipulation, but that’s a different issue.) What would be interesting is a better measure: Was 1999 hotter than the proxies for any previous individual year? The answers, I wager, would be consistently “no.”
    We show smoothed graphs for prior times as if things changed only slowly until industrial Man came on the scene. But the reality is accessible to the scientists, and the results portrayed more fairly would be quite illuminating. I expect that either:
    — the proxy is so completely noisy that it invalidates itself as a proxy, or
    — the proxy’s natural variations will be similar to the natural variations we see now.
    A few examples where annual (or better) data demonstrate this, though I have not done exhaustive work on the idea.
    ===|==============/ Keith DeHavelle

  46. Reply to Keith DeHavelle ==> You have a good point there — Why average when you have actual data points?
    Thanks for the suggestion, I might put something together, drawing from Marcia Glaze Wyatt and Robert Brown on the topic.
    I wrote an essay here on Error Bars that covers a bit of the averaging problem — the fact that an average of a measurement which itself is really a ‘range’ — e.g., 71°F +/- 0.5° (the true description of US temperature records from the glass thermometer era) — must carry forward the same error bar — yearly average 56°F +/- 0.5°.
    My recent essay on “Imaginary” numbers touches on averages as well.

  47. An excellent report. However I think you should look deeper into your own examples. Why is pneumonia included in with flu? Pneumonia is a killer, flu not so much, that is why they lump them together so the flu sounds more scary. I think if you separate the two you will find that flu only directly accounts for about 100 deaths! Not scary enough to convince people to get a flu shot eh.

    • Reply to Stephen ==> Perfect example of asking the first question” What exactly are they really counting?
      One must always be careful when imputing “reasons” for the actions of others. The coding system used by coroners to assign “Cause of Death” is called ICD10. The ICD10 list of causes of death document is 89 pages. “Death by Influenza” is one code, and “Death by Pneumonia” is another code.
      Influenza is defined as “any of several acute highly contagious respiratory diseases caused by strains of three major orthomyxoviruses “. Pneumonia is defined “a disease of the lungs that is characterized especially by inflammation and consolidation of lung tissue followed by resolution, is accompanied by fever, chills, cough, and difficulty in breathing, and is caused chiefly by infection.”
      People die of influenza “when certain complications develop; usually these are pneumonia and worsening of the existing conditions.”
      Thus many coroners, by the time they see a dead body, can tell the person died of pneumonia (which is a symptom or condition) which could have been caused by influenza or some other infection which centered in the lungs, indistinguishable without a test for specific viruses.
      I suspect that this explains the “lumping” of flu and pneumonia deaths in this case.

  48. Few media stories have been more misleading than the series of reports of recovery of Arctic sea ice extent. After each record low, stories by Delingpole, Booker and others report a “recovery”. There was a record low in 2007, so in 2008 and 2010 we get articles scoffing at those claiming ice is reducing. No, they say, it is now recovering or returning to average after the record low of 2007. Totally ignoring the fact that they are describing inter-annual changes not long term trends. After a record low it is likely that subsequent years will return to the long term trend – that is they will be higher than the record low. Still, they would not make that mistake twice, would they?
    Oh look! A new record low in 2012. So what do we get? Articles since then saying ice has recovered since the record low in 2012! It is quite staggering really. All long term trends are ignored, summer ice compared to winter ice, data cherry picked, anecdote used as evidence, factual errors made.
    We need to look at what they are measuring – or what measurements they are reporting. If claims are made about trends (recovering) then measuremnts need to be consistent and over extended periods.

    • (Note: “Michael Darby” is the latest fake screen name for ‘David Socrates’, ‘Brian G Valentine’, ‘Buster Brown’, ‘Joel D. Jackson’, ‘beckleybud’, ‘Edward Richardson’, ‘H Grouse’, and about twenty others. The same person is also an identity thief who has stolen legitimate commenters’ names. All the time and effort he spent on writing 300 comments under the fake “BusterBrown” name, many of them quite long, are wasted because I am deleting them wholesale. ~mod.)

    • Reply to seaice1 and Michael Darby ==> Yes — “What exactly are they really counting?” and ” Is the thing they counted really a measure of the thing being reported?”
      Is “Arctic sea ice extent” really a measure of the thing being reported? — which is usually taken by the general pubic to be “How much sea ice is there in the arctic?” Sea Ice specialists know and acknowledge that these are not the same thing — only related. What they mean to communicate, when they say “Arctic Sea Ice at record low”, is usually along the lines of “The Arctic is getting so warm that the ice is melting, all due to Global Warming/Climate Change.” They generally have not measured the amount of sea ice and usually not the air temperatures year around nor the sea water temperatures — which would tell us if the Arctic is suffering from warming.
      The Danes do keep a record, which generally shows that the Arctic is less frighteningly cold in the Winters over the last decade or so…but that Summer temperatures are unchanged.

      • Kip Hansen. We must of course differentiate between the science publications and the media reports. I think the scientific data is very clear. When they say “sea ice extent” is is defined well and not open to much interpretaion. It is clear that the sea ice minimum extent has been on a downward trend for decades and shows no sign of changing.
        If media reports “Sea ice extent at record low” we do not know what they are intending to communicate. In the specific examples of Delingpole and Booker we do know because they tell us. For example, that increase in sea ice extent from 2007 to 2010 has made fools of alarmists. There may also be examples of media reports exagerating the significance of reduced sea ice, but that is different from denying the trend in sea ice that exists.

        • seaice says:
          …“sea ice extent” is is defined well and not open to much interpretaion.
          I’d call ±15% pretty vague, myself.
          And there has never been any empirical evidence produced showing that human CO2 emissions cause Arctic ice to decline. But that is clearly the message.
          If Arctic ice disappeared it would be a net benefit, with no downside. But as with all climate alarms, the slight drop in Arctic ice (which as since reversed) is claimed to be a sign of approaching DOO-O-O-O-O-M!
          Repent, climate sinners! The end is nigh! ☺

      • (Note: “Michael Darby” is the latest fake screen name for ‘David Socrates’, ‘Brian G Valentine’, ‘Buster Brown’, ‘Joel D. Jackson’, ‘beckleybud’, ‘Edward Richardson’, ‘H Grouse’, and about twenty others. The same person is also an identity thief who has stolen legitimate commenters’ names. All the time and effort he spent on writing 300 comments under the fake “BusterBrown” name, many of them quite long, are wasted because I am deleting them wholesale. ~mod.)

      • Reply to this Arctic Sea Ice Comment Thread ==> My point is simply that “Arctic Sea Ice Extent” is not the same as “Amount of Arctic Sea Ice” — they are two different metrics which measure different things, different aspects of the conditions of Arctic Sea Ice — one measured in square kilometers, one measured in million metric tons (or something like that — a measurement of mass).
        When we ask the second question of my essay, “Is the thing they counted really a measure of the thing being reported?” the answer might be different for each metric, depending on the claim made for the significance of the metric as presented in a scientific paper or a news report.

      • Kip Hansen. I agree – the specific claim should be made clear. If we are talking about extent we should make this clear. If we are talking amount of ice, that is a diifferent metric. Often one will be related to the other, although not identical. This I believe is what leads to erroneous and lazy reports in the media.
        Media reports often take liberties with science to the detriment of public understanding. This occurs in all areas. I was pointing out that one of the most egregious examples of this was in the reporting of sea ice recovery In this case what is measured – annual extent – is used to male a claim about long term trends.
        To some extent we must expect dire reporting from the media. What is less forgivable is press releases about journal papers that go well beyond the claims made in the paper itself.
        dbstealey “I’d call ±15% pretty vague, myself.”
        I am not sure wht you mean here, but perhaops you are refering to the sea ice extent 15% figure often quoted. The sea ice extent is not quoted to +/-15%. At the edge of the shelf the ice breaks up and we get water / ice mixed in. We go from 100% ice to 100% water over some distance. At some point we must decide where the “edge” of the ice is. It doesn’t matter too much exactly what figure you pick, as long as you are consistent and not at the extremes. The usual figure is where ice is 15% of the surface. There is also some interpretation required at coastal zones, where pixels may overlap land and sea. That is why I said “not much” interpretation. There is a little room for interpretation, but really very little compared to estimating for example ice volume, where thickness must be estimated as well as area. The sea ice extent is about as close as we get to a measurement in climate science as it is taken from counting pixels.

Comments are closed.