Fix Your Statistics

Quote of the Day

Some simple things that could be done to reform Earth science from its present disaster state:

– Discard the idea of ‘p-values’ and ‘statistically significant’ data. Report likelihood functions and as much raw data as possible; have an epistemology in which different effect sizes are different hypotheses (since they make the data differently likely). A large experiment that suggests an effect size of 0.05 has *failed* to replicate a small experiment with an effect size of 0.2.

– This will also make “meta-analyses” much easier; point up when different hidden experimental conditions are producing different effect sizes and naively implying incompatible states of reality; reduce positive bias in reporting experimental results by averting the idea that an ‘insignificant’ result is a failure; etc etc. A lot of the problems in modern science really are just the result of using epistemically awful math!

Eliezer Yudkowsky @ESYudkowsky
Tweeted on April 24, 2023

This applies to more than statistics for Earth science.

Of course, I could be biased. I like working with likelihood ratios/functions.

Read the whole thing for suggestions beyond just the math.


8 thoughts on “Fix Your Statistics

  1. I’m not sure what “earth science” means. It sounds like geology, but given the thing about statistics abuse I suspect the intended meaning is climatology. That used to be a science, but it has been destroyed by left wing politics and is now a partisan non-science academic subject, just like “gender studies” and things like that.

    • Earth Sci is studying any of the natural earth systems: weather and climate, geography and geology, live environmental systems and their relation to the inorganic components of the natural world.

      Personally, I’m coming to see a LOT of things as being probabilistic in nature, even things like history, as there are a lot of unknown and hidden variables.

      • My favorite thing to do lately, in connection with climate and “global warming” discussions, is to point people at the GISP2 data set. That’s a record of 40k years of temperature as captured by the Greenland ice sheet.

        Among the interesting points is that the reference point used by the warmists for “pre-industrial climate” is 1850, give or take. That also happens to be the coldest point in the past 8000 years. Coincidence? I doubt it.

        Another interesting point is that the “the world is ending” temperature level pushed by these people is about the climate enjoyed by Julius Caesar, or perhaps a bit cooler. And a few thousand years before him it was a bit warmer still.

        Anthropogenic global warming? I don’t believe it, not until I see a plausible explanation for the temperature variations in the GISP2 data, and a good argument for why the mechanism that caused those isn’t the one in play with the modest temperature rise seen in the past 100 years.

        • I like the “residence time” argument.
          What’s the residence time for CO2 in the atmosphere? That is, how long on average does a molecule of CO2 remain in the atmosphere, from the time it’s emitted to absorbed by a plant or otherwise sequestered from the air? It’s it’s significantly shorter to equilibrium than the observed temperature trend, the only way it can rise is if the rise faster than emissions is if it’s part of a natural cycle.

          For one, we don’t really know for sure; there IS no consensus. Secondly, the range now used by climate scientists are almost all between 5 and 30 years. How can something that will return to equilibrium in, say 10 years, cause a 150 year uptrend unless either massively more CO2 is added each year, or there is an underlying cycle that is pushing the equilibrium point for CO2 entering and leaving the atmosphere upwards? Since we are not adding hugely more each year on that scale, then there must be an underlying trend; that means our CO2 emissions don’t really make a damn bit of difference one way or another.

          • What’s the resident time for human made CO2, is the real question. Since our human share is only 5% or less of CO2 produced in the world. Around 20 PPM out of somewhere around 400 PPM total.
            Nothing we could do would change anything ever in that regard.
            On top of the fact that plants use it for food. I believe it’s Charles law that shows 1 liter of water at 20C and 1 atmosphere of pressure will absorb 1 liter of CO2.
            The higher the pressure, colder the water, the more it will absorb. So freezing temps and high pressures at 10,000ft. deep oceans are nowhere full. (think club soda).
            So every time a wave curls and breaks, or it rains. CO2 is being sequestered. Ocean currents and temperature have the biggest effect. Not our 5%.
            My chemistry book from 1955 postulates that the oceans are one of the biggest reasons CO2 stays somewhat constant in the atmosphere.

  2. Like Pkoning, I’ll assume were talking climate science. Reforming it is one of the easiest things to do. Take the narrative off the table, then quit paying for it.
    Like so many other things that trouble our society today. There being pushed for a reason other than there stated cause.
    And almost all of the statistics generated are for the real object. Money, power, control.
    I’m sorry, but anyone telling you CO2 is a pollutant or poison, or that both that and methane builds up in the atmosphere. Is too ignorant to have an opinion worth listening to. Regardless of education or financial status.
    And anyone working toward that basis is a fraud.
    If Alex Jones said the government was controlling the weather. (As I’m sure he has.) He’s a dangerous nut job! Is all we hear from science.
    When John Kerry and Big Al jet set all over the world demanding you give them money, power, and control over your lives so they can what. Control the weather????? Ya, and you’ll need to eat bugs to!
    If you want accurate statistics. Try using raw data to disprove lies. Not establish them. Then quit changing the measurement protocols/models, to fit your narrative.
    Right now the experts of the world in all classes are busy destroying themself and their credibility. It will not end well for anyone.

    • Part of the problem with people pushing “CO2 is a pollutant” is that often they aren’t ignorant, they are malicious. By thinking of them as merely misinformed we’re giving them too much credit.

    • At this point, I think the raw data has been corrupted too. It would get rid of the “adjustments” which are always in one direction. In theory, adjustments are legitimate in dealing with data sets but in practice, in this case, they are abusive. One problem with the raw data is the vast expansion of urban areas, putting the data collection points within the urban heat island whereas before many of them were outside of it. Another is the unreliability of recording in many areas of the world and the recent provenance of the data. Satellite data is more accurate but of even more recent vintage.
      Whatever you say about climate, it is about long cycles not short term effects. There have been attempts to get around this using tree rings, ocean deposits, ice cores and the like but these have largely been done by the same people that corrupted the current data.

Comments are closed.