The Five Types of Null Hypothesis Error
Defending the integrity of the methods of science is not the same thing as tendering conclusions in lieu of science. Ethical Skeptics get this, fake skeptics do not. An additional litmus test with respect to being able to tell a fake skeptic from a real one, can be found in how they regard disposition of a null hypothesis inside hypothesis reduction theory. Do they conflate statistical hypothesis reduction with scientific hypothesis reduction? Do they employ the incorrect null hypothesis? One might be surprised to find that even in the disciplined halls of institutional science, abuse of the null hypothesis is one of the most common forms of pseudoscience.
The basic definition of the null hypothesis (H0) is as the preeminent and referential member of the mutually exclusive set of options to a sponsored alternative hypothesis (H1). In statistical hypothesis testing and often experimental testing, they are opposites. In scientific hypothesis reduction, this is not necessarily the case; often the two residing only as mutually exclusive ideas, and not necessarily representative of the entire domain of potentiality. The alternative hypothesis is the contention which is being tested for merit, which could serve to falsify the null hypothesis. The null hypothesis is the contention which is assumed to be the standard against which the alternative is to be evaluated. The null however is not assumed to be true in a hypothesis reduction (series of hypothesis tests and eliminations), neither in advance of nor after testing completion in which an alternative fails. The null hypothesis is a hypothesis which the researcher tries to disprove, reject or nullify. This misconception of the null hypothesis standing as the currently ‘true’ hypothesis stems from its oft framing as the the most familiar and accepted view of a contention, or
- the common view of something,
- the accepted view of something,
- the historical view of something.
Often as well, in the instance when a common view of something is not held, reduction may be assembled in terms of a null hypothesis which is
- the parsimonious alternative (least complicated, predicate dependent or feature-stacked).
None of this qualification for the null hypothesis amounts to anything which would pass a Popper Demarcation test for falsification based science by any means. Which introduces the circumstance when, despite the array of predictive and suggestive evidence (or Promotion), when the predictive alternative cannot ethically be promoted to status as the null hypothesis. Specifically, when the ramifications of a hypothesis dictate conservancy in its application; say, regarding an idea’s safety or potential disruption of science or society for instance, the null hypothesis is that which defends such risk, or
- the alternative which introduces the least external risk in its application.
In the past in my labs, when introducing a new material for industrial component fabrication or compounding, even though the constituents of the altered material might be commonly assumed to be safe or might present the most parsimonious view of complexity with respect to the solution it entailed, we would default to the null hypothesis citing that the compound is potentially human health impacting until rigorously proved otherwise (see Error of the Guilty Null below). Many of our products never made it to market because we could not answer this question sufficiently. The guilty null ruled in that case.
Food science ethically is practiced under the Guilty Null ethic. Food is guilty until proved innocent. This may be a pain in the ass, might piss off stockholders and irritate fake academic payroll skeptics and impatient executives. But presuming to tamper with the substances that all of us consume, 3 times a day for life, demands a much higher level of scientific rigor than many social epistemologists are willing to tolerate. Certainly more than the 3 years of study it took to approve glyphosate, for instance.
Under this null hypothesis ethic, we would then seek to nullify the idea that the material presented a health or well being danger, in both short and long term use. This null hypothesis framing can also be seen as a form of parsimony itself; as risk, is indeed also a form of complicated-ness (not complexity). This introduces the principle of the abuse of the null hypothesis as a means to enact ‘truth’ which has not in reality earned this position of merit. Specifically, five types of error are committed by those who are eager to promote an idea through manipulation of the null hypothesis. The first two are formal and the final three are procedural. All five are fallacies of philosophy and science.
A. Null Hypothesis Formal Fallacy
- Argument from Ignorance (Extrapolation from False Positive)
- Proof by Negative Composition (Extrapolation from a False Negative)
B. The Omega Hypothesis Procedural Fallacy
- Error of the Default Null
- Error of the True Null
- Error of the Guilty Null (The Precautionary Principle)
Through the art of the Elegant Lie, I can manipulate thought so as to enforce an entire religion. That is to say, by promotion of unmerited thought into positions of defended science, one can prematurely convince entire masses of people, that questionable thinking is indeed proved science.
A. Null Hypothesis Formal Fallacies: Conflating Statistical/Experimental and Theoretical Hypothesis Framing
In general, an experiment conclusion always refers to the null, rejecting or accepting H0 rather than H1. The null hypothesis stands in the crucible. In statistical analyses however, often we do test a circumstance where H0 is indeed constrained to the outcomes true or not true. The conflation of statistical theory inside diagnostic hypothesis evaluation, with the broader set of science hypothesis reduction, often lends to confusion in broader hypothesis reduction practice. Not every experiment can be evaluated through the employment of p-value delimited distribution curves; nor can a μ0 = μ1 circumstance always be identified. One such example can be seen here with respect to the understanding and employment of type I and II errors inappropriately applied outside the constrained context of a statistical test.
Wikipedia cites regarding type I and type II null hypothesis testing errors:
In statistical hypothesis testing, a type I error is the incorrect rejection of a true null hypothesis (a “false positive”), while a type II error is the failure to reject a false null hypothesis (a “false negative”).¹
This principle is incorrect however, when being applied to theoretical hypotheses.
In broader theoretical hypothesis reduction a false positive and false negative impart no disposition whatsoever to the veracity of the null hypothesis; only the signal which was measured regarding the alternative hypothesis. Confusing the domain of statistical hypothesis testing with the larger practice of scientific hypothesis reduction is a principal tactic of social epistemologists. In logic, as opposed to statistics, the null hypothesis remains in a moot disposition under both a false positive and a false negative condition – and is not therefore assumed true – in scientific hypothesis reduction. In science, as apposed to constrained statistic set theory, when a type I or II error is introduced, a further disposition of the null hypothesis must be separately indicated. If such type I and II error theory is to be correctly framed inside the context of scientific testing, it should read:
In scientific hypothesis testing, a false positive outcome is the incorrect rejection of the null hypothesis which further can lead to a type I error, the incorrect rejection of a true null hypothesis. While a false negative outcome is the incorrect elimination of an alternative which can further lead to a type II error, the failure to reject a false null hypothesis.
Null hypothesis error type and separately an alternative hypothesis’s status as false negative or positive are not the same thing. To force the two outcome dispositions to indeed constitute the same logical equivalent of a type I or type II error constitutes exercise of decision theory inside a bifurcation fallacy. I have observed a cadre of clinical neurologists speak often of examples of type I and II errors, inexpertly identifying cases wherein the two options are not opposites. Ideally, when constructing a critical path for experimentation, one should seek to establish a reductive/deductive series of opposite Bayesian inference tests in which a posterior inference can be drawn. But this is less often the reality. Most errors in scientific testing result relate to controls or constraints (CC), measure (M) or significance (S) errors. A great summation of this principle from even a statistical testing standpoint can be found here: Never Found a type I or II error (Science Modeling, Casual Inference and Social Science).
This principle, the abuse of type I and II error contexts, is displayed in the graphic above, where such a bifurcation is shown to be enacted through an implicit argument from ignorance or fallacy of negative composition. Categorization of a hypothesis testing, outside of Bayesian theory, to have constituted a type I or type II error is a disposition that can be assigned only after the hypothesis reduction has been completed in finality. Never before. But you will find social epistemologists swinging the term inside pluralistic arguments like they were an erstwhile major league batter of science.
A principle method of deception-by-pretense employed by social epistemologists (Social Skeptics) is the prejudicial framing of arguments as constituting type I and type II errors, while indeed the science is still ongoing inside a hypothesis reduction. This is pseudoscience. Be very wary of a Social Skeptic who over-employs the type I and type II error disposition in social discourse.
Type I and II Error Abuse Fallacies
By forcing a false positive or negative, or spinning alternative outcomes inside a diagnostic or statistical ‘true or not true‘ theory basis, to therefore imply without sufficient evidence a conclusive hypothesis reduction, of the nature of a type I or II error, I have committed:
Argument from Ignorance
/logic : fallacy of argument/ : asserts that a proposition is true because it has not yet been proven false, or is false because it has not been shown to have any evidence.
Science has found no proof of intelligent life nearby us in space, therefore intelligent life does not exist nearby us in space.
Science, despite its best efforts, cannot disprove the concept of existence of a god, therefore God exists.
Proof by Negative Composition
/logic : fallacy of argument/ : disproof of tenets inside an opponent’s idea or of the idea itself stands as proof of my own idea or argument.
Discrediting of the Piltdown Man fossilized remains as a paleoanthropological hoax, demonstrates that evolution is a fraud and stands as a proof of the validity of Creation Theory.
B. The Omega Hypothesis: Abuse of the Null Hypothesis
An additional form of pseudoscience which fails the Popper Demarcation principle, can be found in the practices regarding the employment of an invalid null hypothesis, HΩ. The creation and unmerited protection of the Omega Hypothesis constitutes a form of hypoepistemology which is spun through practices of Inverse negation fallacy, and corruption of the standards and methods of science. Through these practices of social epistemology, an apparent coherence can be spun around a particular view of a subject, and protection by the corrupted institutions of science afforded until such time as a Kuhn Paradigm Shift is able to be precipitated. Sadly, this often only occurs upon the death of the key social epistemologists involved.
Omega Hypothesis (HΩ)
/philosophy : pseudoscience : social epistemology : apparent coherency/ : the argument which is foisted to end all argument, period. A conclusion which has become more important to protect, than the integrity of science itself. An invalid null hypothesis or a preferred idea inside a social epistemology. A hypothesis which is defined to end deliberation without due scientific rigor, alternative study consensus or is afforded unmerited protection or assignment as the null. The surreptitiously held and promoted idea or the hypothesis protected by an Inverse Negation Fallacy. Often one which is promoted as true by default, with the knowledge in mind that falsification will be very hard or next to impossible to achieve.
Einfach Mechanism – an explanation, theory or idea which resolves a contention under the scientific method solely by means of the strength of the idea itself. An idea which is not vetted by the rigor of falsification, predictive consilience nor mathematical derivation, rather is simply considered such a strong or Occam’s Razor (sic) simple an idea that the issue is closed as finished science from its proposition and acceptance onward. An einfach mechanism may or may not be existentially true.
Höchste Mechanism – when a position or practice, purported to be of scientific basis, is elevated to such importance that removing the rights of professionals and citizens to dissent, speak, organize or disagree (among other rights) is justified in order to protect the position or the practice inside society.
Entscheiden Mechanism – the pseudoscientific or tyrannical approach of, when faced with epistemology which is heading in an undesired direction, artificially declaring under a condition of praedicate evidentia, the science as ‘settled’.
Skulptur Mechanism – the pseudoscientific method of treating evidence as a work of sculpture. Methodical inverse negation techniques employed to dismiss data, block research, obfuscate science and constrain ideas such that what remains is the conclusion one sought in the first place. A common tactic of those who boast of all their thoughts being ‘evidence based’. The tendency to view a logical razor as a device which is employed to ‘slice off’ unwanted data (evidence sculpting tool), rather than as a cutting tool (pharmacist’s cutting and partitioning razor) which divides philosophically valid and relevant constructs from their converse.
Verdrängung Mechanism – the level of control and idea displacement achieved through skillful employment of the duality between pluralistic ignorance and the Lindy Effect. The longer a control-minded group can sustain an Omega Hypothesis perception by means of the tactics and power protocols of proactive pluralistic ignorance, the greater future acceptability and lifespan that idea will possess. As well, the harder it will to be dethrone as an accepted norm or perception as a ‘proved’ null hypothesis.
One key sign that an Omega Hypothesis is being promoted, is the tactic of declaring any non-conventional alternative explanation as constituting ‘magical thinking.’ This paranoia about every thought that threatens one’s beliefs as stemming somehow from magic, is in itself a version of religious thinking. Three conditions typically lead to this tactic, and highlight a person’s religious clinging to the Omega Hypothesis:
A. Forcing a Null Hypothesis from an idea which has not really been matured into an actual scientific hypothesis in the first place,
B. Assuming the Null Hypothesis to be true,
C. Assuming all competing hypotheses to be declarations of ‘magical thinking’ – in an attempt to obviate any scientific testing or maturing of such an idea.
These stand as the warning signs that a social epistemology of bullying, and enforcement of the Omega Hypothesis, is in play.
All the above a set of practice which abrogates a Popperian view of the threshold and rigor of adequate science, relying instead on the promotion of an invalid null hypothesis (HΩ) through academic inertia, ignorance of the discipline, promotification science, social skeptic campaigns or corporate pressure.
Assignment of the incorrect null is in reality a form of asking the wrong or begged question under the scientific method. I can prejudice the results of science by tampering with that which is assumed as its null. Specifically, this entails several forms of favoring a null hypothesis by assigning it unmerited status as the null hypothesis through a series of non discriminating, but sciency-looking predictive tests, – promotification or King of the Hill science practices – and moreover, through ‘parsimony’ or ‘Occam’s Razor’ default, granting the incorrect null unmerited status as generally accepted scientific theory until such time as the monumental task of disproving it, is achieved. Karl Popper cited in his essay on pseudoscience, Science: Conjectures and Refutations, element (4) in his list of 7 things which distinguish science from non-science:
(4) A theory which is not refutable by any conceivable event is nonscientific. Irrefutability is not a virtue of a theory (as people often think) but a vice.²
This is the essence of the first of the two null hypothesis assignment errors. The Error of the Default or Irrefutable Null. There are two forms of such tampering with the null hypothesis:
Error of the Default Null
/philosophy : pseudoscience : incorrect hypothesis testing/ : a variation of argument from ignorance.
The practice of assigning a favored or untestable/unfalsifiable hypothesis unmerited status as the null hypothesis. Further then proclaiming the Default Null as the null hypothesis until such time as it can be defeated by new competing science.
Since science has conducted no research into the possible existence of a spiritual realm, therefore the spiritual realm exists, stands as the null hypothesis until such time as this can be disproved by science.
Running promotification tests supporting an idea when test data falsifying that idea already exists.
I possess an accurate definition of the term God.
You cannot prove that god does not exist, therefore God exists until otherwise proved
You cannot measure God, therefore no such thing as anything unmeasureable exists.
You cannot prove that god does not exist, therefore God exists until otherwise proved
I possess credible science and resources regarding representation of scientific consensus.
Error of the True Null
/philosophy : pseudoscience : incorrect hypothesis testing/ : a variation of argument from ignorance.
Regarding the null hypothesis as objectively ‘true’ until proved otherwise, when it simply is the null hypothesis from the standpoint of the logical calculus in a hypothesis reduction hierarchy and not because it has been underpinned by a Popper level scientific rigor. Further then proclaiming the True Null to be the prevailing conclusion of science.
Consciousness is material monist in origination. Science has shown that all other hypotheses are false.
Science has ‘found no evidence’ regarding the existence of a spiritual realm, therefore no investigation should be conducted until such time that absolute proof is obtained.
A suspect in a very difficult legal case should be considered guilty until such time as he is able to prove he is innocent.
A subject is scientifically considered a pseudo-science until such time as it can be proved real.
You cannot measure God, therefore under science any thing which it has not measured objectively does not exist.
Error of the Guilty Null (Precautionary Principle)
/philosophy : pseudoscience : incorrect hypothesis testing/ : the practice of assigning a favored hypothesis the status as null hypothesis, when in fact the hypothesis involves a feature or implication which would dictate its address as an alternative hypothesis instead. A null hypothesis which is, by risk or impact, considered potentially harmful until proved innocent, should be treated as an alternative under correct parsimony. Further then invalidly proclaiming this Guilty Null to be the prevailing conclusion of science until such testing is conducted which could prove it to be false or until such time as it can be defeated by new competing science.
A violation of the precautionary principle, in this case a ‘lack of directly associated hazard’ is commuted to mean ‘lack of necessity to observe or measure risk.’ Since risk cannot, has not, or has been shown through scant study to possibly be low, then the substance is approved for use until proved harmful. It is fraudulent pseudoscience. When stats of a substances harm are further squelched, the pseudoscience becomes a crime.
Presuming disease to be simply a matter of bodily system malfunction, rather than to stem from external inducement (environment (involuntary), lifestyle (voluntary), epigenetic, viral, bacterial/biome, genetic) – while logically such a contention is easy to default and could stand as a null hypothesis, such a practice introduces extraordinary amounts of stacked provisional knowledge and risk. It fails the precautionary principle. This again is a common trick of pseudoscience.
A new crop control pesticide is safe for human consumption until proved by science to be harmful/inflammation inducing.
DNA and protein chains from food never make it past the human digestive barrier so therefore polynucleotides and protein chain based hormones added to animal stocks do not impact human endocrine systems.
Light dimming switches are safe for consumers until such time as house fires can be objectively linked beyond correlation, to their introduction into the market.
The standard of testing of safety for a compound which is applied topically or ingested occasionally, should be the same standard applied for a compound which is ingested daily/regularly.
An example of The Guilty Null and The Default Null can be found here with respect o Nassim Nicholas Taleb’s Precautionary Principle (click on image below):
/business : consumer social policy : medicine and food/ : the lifecycle management of chemicals, adjuvants or biological agents which do not indicate immediate classic major pathology pathways in test animals, into a final phase of testing upon the broader human population, in order to speed them to market and generate revenue during long term employment testing. Establishment of activist ‘skeptics’ to patrol and ensure any failures are squelched as constituting only pseudoscience and anecdote.
Keep a sharp eye out for both the type I and II error claims as well as practices of King of the Hill pseudoscience, as such fallacies inside of hypothesis reduction regularly occur inside science as well as the social discourse. Both fallacy sets can masquerade as real science if keen minds are not watching the candy store. In essence, deployment of the two forms of the Omega Hypothesis stands as a way of interceding on behalf of science. By skeptically upstaging science I can therefore pretend to speak on behalf of science by socially corrupting its methods.
Ethical Skepticism Litmus
As an Ethical Skeptic, you are the one tasked with maintaining a discriminating mind with regard to process. No, you are not claiming to represent science or its conclusions. But one can as a skeptic indeed take a stand to defend the method of science when one observes it being abrogated. This is what the term ‘ethics’ means, an allegiance to a standard of practice – and not an allegiance to a particular set of outcomes, or categories of thought one considers socially valid or invalid.
Protecting the integrity of the methods of science is not the same thing as tendering conclusions in lieu of science.
This differentiation is a key litmus test in being able to tell who is a fake and a real skeptic. Real skeptics get this difference. False skeptics do not.
¹ Wikipedia: Type I and type II errors, https://en.wikipedia.org/wiki/Type_I_and_type_II_errors.
² Karl R. Popper: Science: Conjectures and Refutations. Conjectures and Refutations (1963), p. 43–86;
No comments yet.