The Null Hypothesis and Global Warming


The Reference Frame points out an interesting article about the Null Hypothesis and Climate Change.

For those who don’t know, the Null Hypothesis is a way to run experiments and generate meaningful results. It starts with a simple statement that can be easily disproven, the Null Hypothesis. Then you collect your data, and you evaluate what the probability of the Null Hypothesis being true is, given the data.

As an example, consider the Null Hypothesis “this coin is fair”. If you flip the coin 5 times, and you get 5 heads in a row, then you note that the probability of getting a run of 5 heads with a fair coin is 6%, so the Null Hypothesis is only 6% likely to be true.

It’s important to note that you can’t ever really rule out a Null Hypothesis, unless it is stated in terms of absolutes. In our coin example, there is still a tiny change the Null Hypothesis is true, even if you flip the coin a thousand times and get heads every time. However, the chance of the Null Hypothesis being true in such a case is very, very small. Conversely, you can never really show a Null Hypothesis to be true, unless it is stated in terms of absolutes.

What does this have to do with Climate Change or Global Warming?

In the article I cited above, one scientist says, “The Null Hypothesis should be that Global Warming is true.” Another scientist says, “No, it shouldn’t be, and Null Hypothesis are pointless anyway.” A third person, a mathematical physicist says, “You’re both wrong.”, and goes on to say that the Null Hypotheses (plural here–multiple hypothesis-es) that have been proposed in the past have never been shown to be wrong.

Of course, the Reference Frame goes on to point the silliness about the whole thing with basic and clear science and observations.

To date, I haven’t seen many Null Hypothesis that even survive the laugh test from climate science. I certainly haven’t seen many genuine experiments.


6 Responses to “The Null Hypothesis and Global Warming”

  1. True Epicurean Says:

    Your understanding of what certain scientific terms mean seems… well… quite flawed. Would you happen to know where we could get another Earth as a control for an “experiment”?

    Likewise, it seems that you don’t really understand the question these folks are debating. The point is simple: how do you determine the base case for a “null hypothesis”? In your oversimplified thought experiment, you propose that the base case — what you believe before the experiment begins — is that there is a very high probability that the coin is “fair”. That probability goes down with each head that comes up. However, you could easily reverse this question, and ask what the probability is that the coin is “unfair”. The only difference would be the burden of proof — how much evidence would be required before you would switch your opinion (and the null hypothesis would be disproven).

    Now, despite your pained bleatings to the contrary, there is substantive scientific evidence that there is some change in climate, and that change in climate is recent and not likely to be caused by natural phenomena. This sufficient evidence provides a basis for authorities in the field to state that this is occurring. If that is the case — regardless of your political biases that make you prefer otherwise — why shouldn’t the “null hypothesis” change?

    • Jonathan Gardner Says:

      Please, if you have evidence, present it. No one says you shouldn’t.

      However, as I watched with the Greenhouse Effect, something which I was sure people had gotten right, sometimes your most basic assumptions turn out to be flat-out wrong.

      Can you restate your experiments in terms of a falsifiable null hypothesis? Can you define what “natural processes” are, what a “change in climate” means, and what you mean by “recent”? Otherwise, we’ll be chasing your tail until the cows come home. As you’ll quickly find, the principle problem with the climate hypotheses is that they are not quantifiable, the science behind the effects turns out to be chaotic, IE, unpredictable, and you end up with things that can’t be clearly shown.

      As it is, the earth is much, much more massive and expansive than people realize, and humans have a much, much more minor effect than you think. If we tried to poison the oceans by pumping as much arsenic as we use oil into them, it would be hundreds of years before there would be a noticeable change, and arsenic is probably one of the worst poisons we release into the environment. The same goes for the atmosphere—even the worst acid rains were a localized effect, limited only to the offender and its immediate neighborhood, compared to the entire earth. The strongest nuclear bombs we have set off in the atmosphere register only a tiny change in the net radioactivity of the air, a change so small that unless you have advanced tools looking specifically for such a change, you would never see it. We humans are, as a matter of fact, insignificant to the earth.

      And yet, you believe that modifying the concentration of CO2 in the atmosphere, a trace gas with a hardly noticeable affect in the thermodynamic properties of air, that we’ll cause the entire earth to collapse.

  2. Gene Says:

    This is perhaps the most alarming aspect of the anthropogenic global warming position; the failure of its advocates to utilize a proper null hypothesis in their argument. In any other field of study, the null hypothesis is required to be the statement of the opposite of that which is to be proven, and the data must reject the null hypothesis at a 95% confidence level. The global warming advocates reject this fundamental tenet of scientific and statistical validity. The mere existence of global warming has never been subjected to such rigor; in fact the proponents of the global warming hypothesis secretly admit among themselves (see the Climategate emails) that there has been no demonstrable warming since 1998. The datasets claimed to be independent are not; the surface temperature data collection system was never intended to resolve changes in temperature as small as the advocates claim are present; the corrections to these temperature stations are not justifiable in any rational sense. Most of the source data has been lost/hidden/obscured/destroyed.

    If I had manipulated data this way in my graduate studies, I would have been given an ‘F’.

  3. Gene Says:

    True Epicurian illustrates how the global warming proponents have lost their credibility. The frequent and continuing use of ad hominum attacks, and other logical fallacies, suggests an inability to argue their case based on data and logic. Some examples:

    >Your understanding of what certain scientific terms mean seems… well… quite flawed. Would you happen to know where we could get another Earth as a control for an “experiment”?…it seems that you don’t really understand……how do you determine the base case for a “null hypothesis”

    Red herring; base case is not required for a null hypothesis.

    > In your oversimplified thought experiment……the only difference would be the burden of proof — how much evidence would be required before you would switch your opinion (and the null hypothesis would be disproven). …pained bleatings to the contrary… …substantive scientific evidence……[the] change in climate is recent……not likely to be caused by natural phenomena…. This sufficient evidence provides a basis for authorities in the field to state that this is occurring… If that is the case — regardless of your political biases that make you prefer otherwise —…why shouldn’t the “null hypothesis” change? <

    Switching the Burden of Proof fallacy… that's why. That was the point. A proper null hypothesis must be established and tested against reliable data that is statistically valid at a minimum of a 95% confidence level. So far, the proponents of the hypothesis have not done this. The surface data is insufficiently reliable; the corrections applied to them have not been justified; the organizations holding the data sets have steadfastly stonewalled outsiders who might audit the case for the corrections; the areas not covered by temperature sensors are assumed to have temperatures as provided by an interpolation algorithm that has never been validated; the surface temperature data does not agree with the satellite data; the famed 'hockey stick' graph combines concatenates dissimilar data sets; the 'recent' temperature rise shown in those graphs is based on tree ring data from a tiny data set in Siberia; tree rings are not reliable temperature records; parts of the data showing decreases in temperature were deleted; Mann has publicly stated that concatenating dissimilar data sets is unacceptable, and then denied it, the facts not withstanding that this is exactly what he did; audits of the data processing algorithm showed that it produced a 'recent' temperature rise no matter what data was fed to the algorithm, even white noise.

    It goes on and on.

    I will listen to the global warming proponents anytime they return to a valid scientific method. Until then, their credibility is zero.

  4. Curt Hobbs Says:

    Thank You guys for your comments. I am only in the virginity of research, as a psychologist ( ewww. ), and learning the basic tenants for testing research. I had learned of “null hypothesis”, in a probability and statistics class and was reading more on it in “Basics of Social Research:Qualitative and Quantitative Approaches”, which ironically had mentioned something about climate research in the beggining of the book. Now I Am on the section of the different types and aspects of hypothesis and wondered what the null hypothesis was of man made global warming considering the tedious task of proving or disproving a hypothesis. Hahaha, Yes, as I suspected after reading through this unforgivable process of proof, there is no way global warming is proven, there can’t even possibly be enough related data or evidence to formulate a test that would prove such a complex phenomenon. Granted, I am very limited in my knowledge but I did enjoy your post and certainly agree with what you two have posted. Clearly having much more extensive understanding of the scientific process than I do. It is frightening so many people fraudulently came out and made this claim in the name of science. What a shame my book even made some mention of it in the beginning as if a person who didn’t believe in it was a fool ( of course not those words, but that was the implication I got). I will have to go back and review what was said now that I consider what I have learned ! :) Thanks again !!

    • Jonathan Gardner Says:

      One of the things about science is that it’s built for idiots, because we are all idiots. If a so-called scientist wants to show you something that they learned, and they can’t answer simple questions you have that would connect your personal observations to their conclusions, then they aren’t doing science, they’re doing magic, or worse, creating a cult.

      Example: A physicist has to be able to show and demonstrate the phenomena of gravity, and do it in such a way that they can convince you, the listener/viewer/reader of the correctness of their ideas. I’d probably start by dropping something, and measuring its speed as it falls, and show you that its speed increases over time in proportion, at least until it begins to approach its terminal velocity. Then I’d point to you the motion of the planets, and show you how Kepler’s equations describe their position with a high degree of accuracy, and then I’d connect Kepler’s equations to the acceleration of gravity we saw on earth. It’s not a simple task, but it is possible, and it must be done before anyone can claim that they “believe” in gravity as we understand it.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: