One and done statistical studies, based upon a single set of statistical observations (or even worse lacks thereof), are not much more credible in strength than a single observation of Bigfoot or a UFO. The reason, because they have not served to develop the disciplines of true scientific hypothesis. They fail in their duty to address and inform.
As most scientifically minded persons realize, hypothesis is the critical foundation in exercise of the scientific method. It is the entry door which demonstrates the discipline and objectivity of the person asking to promote their case in science. Wikipedia cites the elements of hypothesis in terms of the below five features, as defined by philosophers Theodore Schick and Lewis Vaughn:1
- Testability (involving falsifiability)
- Parsimony (as in the application of “Occam’s razor” (sic), discouraging the postulation of excessive numbers of entities)
- Scope – the apparent application of the hypothesis to multiple cases of phenomena
- Fruitfulness – the prospect that a hypothesis may explain further phenomena in the future
- Conservatism – the degree of “fit” with existing recognized knowledge-systems.
Equivocally, these elements are all somewhat correct, however none of the five elements listed above constitute logical truths of science nor philosophy. They are only correct under certain stipulations. The problem resides in that this renders these elements not useful, and at worst destructive in terms of the actual goals of science. They do not bear utility in discerning when fully structured hypothesis is in play, or some reduced set thereof. For instance, ‘Scope’ is functionally moot at the point of hypothesis, because in the structure of Intelligence, the domain of observation has already been established – it had to have been established, otherwise you could not develop the hypothesis from any form of intelligence to begin with.2 3 To address scope again at the hypothesis stage is to further tamper with the hypothesis without sound basis. Let the domain of observation stand, as it was observed – science does not advance when observations are artificially fitted into scope buckets (see two excellent examples of this form of pseudoscience in action, with Examples A and B below).
Fruitfulness can mean ‘producing that which causes our paradigm to earn me more tenure or money’ or ‘consistent with subjects I favor and disdain’ or finally and worse, ‘is able to explain everything I want explained’. Predictive strength, or even testable mechanism, are much stronger and less equivocal elements of hypothesis. So, these two features of hypothesis defined by Schick and Vaughn are useless to vacuous in terms of real contribution to scientific study. These two bad philosophies of science (social skepticism) serve to produce inevitably a fallacy called explanitude. A condition wherein the hypothesis is considered stronger the more select historical observations it serves to explain and how flexible it can be in predicting or explaining select future observations. Under ethical skepticism, this qualification of an alternative or especially null hypothesis is a false notion. Also known as pseudo-theory, an idea which explains everything easily, likely explains nothing at all. This process begins by a faulty method of science which ‘begins with a question’ (aka as a ‘rhetorically-expressed’ a priori answer).
/philosophy : pseudoscience : sciebam/ : a question, purported to be the beginning of the scientific method, which is asked in the blind, without sufficient intelligence gathering or preparation research, and is as a result highly vulnerable to being manipulated or posed by means of agency. The likelihood of a scientifically valid answer being developed from this question process, is very low. However, an answer of some kind can almost always be developed – and is often spun by its agency as ‘science’. This form of question, while not always pseudoscience, is a part of a modified process of science called sciebam. It should only be asked when there truly is no base of intelligence or body of information regarding a subject. A condition which is rare.
/philosophy : science : method : sciebam/ : (Latin: I knew) An alternative form of knowledge development, which mandates that science begins with the orphan/non-informed step of ‘ask a question’ or ‘state a hypothesis’. A non-scientific process which bypasses the first steps of the scientific method: observation, intelligence development and formulation of necessity. This form of pseudoscience/non-science presents three vulnerabilities:
First it presumes that the researcher possesses substantially all the knowledge or framework they need, lacking only to fill in final minor gaps in understanding. This creates an illusion of knowledge effect on the part of the extended domain of researchers. As each bit of provisional knowledge is then codified as certain knowledge based upon prior confidence. Science can only progress thereafter through a series of shattering paradigm shifts.
Second, it renders science vulnerable to the possibility that, if the hypothesis, framework or context itself is unacceptable at the very start, then its researcher therefore is necessarily conducting pseudoscience. This no matter the results, nor how skillfully and expertly they may apply the methods of science. And since the hypothesis is now a pseudoscience, no observation, intelligence development or formulation of necessity are therefore warranted. The subject is now closed/embargoed by means of circular appeal to authority.
Finally, the question asked at the beginning of a process of inquiry can often prejudice the direction and efficacy of that inquiry. A premature or poorly developed question, and especially one asked under the influence of agency (not simply bias) – and in absence of sufficient observation and intelligence – can most often result quickly in a premature or poorly induced answer.
Science – ‘I learn’ = using deduction and inductive consilience to infer a novel critical path understanding
Sciebam – ‘I knew’ = using abduction, panduction and linear/statistical induction to enforce an existing or orphan interpretation
Ethical skepticism proposes a different way of lensing the above elements. Under this philosophy of hypothesis development, I cannot make any implication of the ilk that ‘I knew’ the potential answer a priori. Such implication biases both the question asked, as well as the processes of inference employed. Rather, hypothesis development under ethical skepticism involves structure which is developed around the facets of Intelligence, Mechanism and Wittgenstein Definition/Domain. A hypothesis is neither a hunch, assumption, suspicion nor idea. Rather it is a form of self-skeptical notion:
/philosophy : skepticism : scientific method/ : a disciplined and structured incremental risk in inquiry, relying upon the co-developed necessity of mechanism and intelligence. A hypothesis necessarily features seven key elements which serve to distinguish it from non-science or pseudoscience.
The Seven Elements of Hypothesis
1. Construct based upon necessity. A construct is a disciplined ‘spark’ (scintilla) of an idea, on the part of a researcher or type I, II or III sponsor, educated in the field in question and experienced in its field work. Once a certain amount of intelligence has been developed, as well as definition of causal mechanism which can eventually be tested (hopefully) under a given risk exposure or sufficient plausibility, then the construct becomes ‘necessary’ (i.e. passes Ockham’s Razor). See The Necessary Alternative. A hypothesis is not simply a ‘question’, especially one which is asked through agency, or because the scientific method supposedly ‘starts with a question’.
2. Wittgenstein definition and defined domain. A disciplined, exacting, consistent, conforming definition need be developed as premise for both the domain of observation, as well as the underpinning terminology and concepts. See Wittgenstein Error and The Tests of Neologism.
3. Parsimony. The resistance to expand explanatory plurality or descriptive-feature complexity beyond what is absolutely necessary, combined with the wisdom to know when to do so. Conjecture along an incremental and critical path of syllogism/risk. Avoidance of unnecessarily orphan questions, even if apparently incremental in the offing. See The Real Ockham’s Razor. Three characteristic traits highlight hypothesis which has been adeptly posed inside parsimony.
a. Is incremental and critical path in its construct – the incremental conjecture should be a reasoned, single stack and critical path new construct. Constructs should follow prior art inside the hypothesis (not necessarily science as a whole), and seek an answer which serves to reduce the entropy of knowledge.
b. Methodically conserves risk in its conjecture – no question may be posed without risk. Risk is the essence of hypothesis. A hypothesis, once incremental in conjecture, should be developed along a critical path which minimizes risk in this conjecture by mechanism and/or intelligence, addressing each point of risk in increasing magnitude or stack magnitude.
c. Posed so as to minimize stakeholder risk – (i.e. precautionary principle) – a hypothesis should not be posed which suggests that a state of unknown regarding risk to impacted stakeholders is acceptable as central aspect of its ongoing construct critical path. Such risk must be addressed first in critical path as a part of 3. a. above.
4. Duty to Reduce Address and Inform. A critical element and aspect of parsimony regarding a scientific hypothesis. The duty of such a hypothesis to expose and address in its syllogism, all known prior art in terms of both analytical intelligence obtained or direct study mechanisms and knowledge. If information associated with a study hypothesis is unknown, it should be simply mentioned in the study discussion. However, if countermanding information is known or a key assumption of the hypothesis appears magical, the structure of the hypothesis itself must both inform of its presence and as well address its impact. See Methodical Deescalation and The Warning Signs of Stacked Provisional Knowledge.
Unless a hypothesis offers up its magical assumption for direct testing, it is not truly a scientific hypothesis. Nor can its conjecture stand as knowledge.
/philosophy : pseudoscience/ : A pseudo-hypothesis explains everything, anything and nothing, all at the same time.
A pseudo-hypothesis fails in its duty to reduce, address or inform. A pseudo-hypothesis states a conclusion and hides its critical path risk (magical assumption) inside its set of prior art and predicate structure. A hypothesis on the other hand reduces its sets of prior art, evidence and conjecture and makes them manifest. It then addresses critical path issues and tests its risk (magical assumption) as part of its very conjecture accountability. A hypothesis reduces, exposes and puts its magical assertion on trial. A pseudo-hypothesis hides its magical assumptions woven into its epistemology and places nothing at risk thereafter. A hypothesis is not a pseudo-hypothesis as long as it is ferreting out its magical assumptions and placing them into the crucible of accountability. Once this process is ceased, the ‘hypothesis’ has been transformed into an Omega Hypothesis. Understanding this difference is key to scientific literacy.
Grant me one hidden miracle and I can explain everything else.
5. Intelligence. Data is denatured into information, and information is transmuted into intelligence. Inside decision theory and clandestine operation practices, intelligence is the first level of illuminating construct upon which one can make a decision. The data underpinning the intelligence should necessarily be probative and not simply reliable. Intelligence skills combine a healthy skepticism towards human agency, along with an ability to adeptly handle asymmetry, recognize probative data, assemble patterns, increase the reliability of incremental conjecture and pursue a sequitur, salient and risk mitigating pathway of syllogism. See The Role of Intelligence Inside Science. If all the intelligence offered is cherry picked, mocked, or otherwise biased toward or against the hypothesis, it is not really a hypothesis.
6. Mechanism. Every effect in the universe is subject to cause. Such cause may be mired in complexity or agency; nonetheless, reducing a scientific study into its components and then identifying underlying mechanisms of cause to effect – is the essence of science. A pathway from which cause yields effect, which can be quantified, measured and evaluated (many times by controlled test) – is called mechanism. See Reduction: A Bias for Understanding.
7. Exposure to Accountability. This is not peer review. While during the development phase, a period of time certainly must exist in which a hypothesis is held proprietary so that it can mature – and indeed fake skeptics seek to intervene before a hypothesis can mature and eliminate it via ‘Occam’s Razor’ (sic) so that it cannot be researched. Nonetheless, a hypothesis must be crafted such that its elements 1 – 6 above can be held to the light of accountability, by 1. skepticism (so as to filter out sciebam and fake method) which seeks to improve the strength of hypothesis (this is an ‘ally’ process and not peer review), and 2. stakeholders who are impacted or exposed to its risk. Hypothesis which imparts stakeholder risk, which is held inside proprietary cathedrals of authority – is not science, rather oppression by court definition.
It is developed from a construct – which is a type of educated guess (‘scintilla’ in the chart below). One popular method of pseudoscience is to bypass the early to mid disciplines of hypothesis and skip right from data analysis to accepted proof. This is no different ethically, from skipping right from a blurry photo of Blobsquatch, to conjecture that such cryptic beings are real and that they inhabit all of North America. It is simply a pattern in some data. However, in this case, blurry data which happened to fit or support a social narrative.
A hypothesis reduces, exposes and puts its magical assertion on trial.
A pseudo-hypothesis hides its magical assumptions woven into its epistemology and places nothing at risk thereafter.
Another method of accomplishing inference without due regard to science, is to skip past falsifying or countermanding information and simply ignore it. This is called The Duty to Address and Inform. A hypothesis, as part of its parsimony, cannot be presented in the blind – bereft of any awareness of prior art and evidence. To undertake such promotional activity is a sale job and not science. Why acknowledge depletion of plant food nutrients on the part of modern agriculture, when you have a climate change message to push? Simply ignore that issue and press your hypothesis anyway (see Examples A and B below).
However, before we examine that and other examples of such institutional pseudoscience, let’s first look at what makes for sound scientific hypothesis. Inside ethical skepticism, a hypothesis bears seven critical elements which serve to qualify it as science.
These are the seven elements which qualify whether or not an alternative hypothesis becomes real science. They are numbered in the flow diagram below and split by color into the three discipline streams of Indirect Study (Intelligence), Parsimony and Conservatism (Knowledge Continuity) and Direct Study (Mechanism).
A Few Examples
In the process of defining this philosophical basis over the years, I have reviewed several hundred flawed and agency-compliant scientific studies. Among them existed several key examples, wherein the development of hypothesis was weak to non-existent, yet the conclusion of the study was accepted as ‘finished science’ from its publishing onward.
Most institutional pseudoscience spins its wares under a failure to address and/or inform.
If you are going to accuse your neighbor of killing your cat, if their whereabouts were unknown at the time, then your hypothesis does not have to address such an unknown. Rather merely acknowledge it (inform). However much your neighbor disliked your cat (intelligence), if your neighbor was in the Cayman Islands that week, your hypothesis must necessarily address such mechanism. You cannot ignore that fact simply because it is inconvenient to your inductive/abductive evidence set.
Most all of these studies skip the hypothesis discipline by citing a statistical anomaly (or worse lack thereof), and employing a p-value masquerade as means to bypass the other disciplines of hypothesis and skip right to the peer review and acceptance steps of the scientific method. Examples A and B below fail in their duty to address critical mechanism, while Examples B and C fail in their duty to inform the scientific community of all the information they need, in order to tender peer review. Such studies end at the top left hand side of the graphic above and call the process done, based upon one scant set of statistical observation – in ethical reality not much more credible in strength than a single observation of Bigfoot or a UFO.
Example A – Failure in Duty to Address/Inform on Mechanism, Asking an Orphan Question (Sciebam), Fallacy of Explanitude, Linear Induction used when Deduction was Necessary (Methodical Deescalation)
Increasing CO2 threatens human nutrition. Meyers, Zanobetti, et. al. (Link)
In this study, and in particular Extended Data Table 1, a statistical contrast was drawn between farms located in elevated CO2 regions versus ambient CO2 regions. The contrast resulted in a p-value significance indicating that levels of Iron, Zinc, Protein and Phytate were lower in areas where CO2 concentrations exhibited an elevated profile versus the global ambient average. This study was in essence a statistical anomaly; and while part of science, should never be taken to stand as neither a hypothesis, nor even worse a conclusion – as is indicated in the social skeptic ear-tickling and sensationalist headline title of the study ‘Increasing CO2 threatens human nutrition’. The study has not even passed the observation step of science (see The Elements of Hypothesis graphic above). Who allowed this conclusion to stand inside peer review? There are already myriad studies showing that modern (1995+) industrial farming practices serve to dramatically reduced crop nutrient levels.4 Industrial farms tend to be nearer to heavy CO2 output regions. Why was this not raised inside the study? What has been accomplished here is to merely hand off a critical issue of health risk, for placement into the ‘climate change’ explanitude bucket, rather than its address and potential resolution. It broaches the question, since the authors neither examined the above alternative, nor raised it inside their Discussion section – that they care neither about climate change nor nutrition dilution – viewing both instead as political football means to further their careers. It is not that they have to confirm this existing study direction, however they should at least acknowledge this in their summary of analytics and study limitations. The authors failed in their duty to address standing knowledge about industrial farming nutrient depletion. This would have never made it past my desk. Grade = C (good find, harmful science).
Example B – Failure in Both Duty to Inform of Intelligence and Duty to Address Mechanism, Fallacy of Explanitude, Orphan Question (Sciebam), Linear Induction Employed
Possible future impacts of elevated levels of atmospheric CO2 on human cognitive performance and on the design and operation of ventilation systems in buildings. Lowe, Heubner, et. al. (Link)
This study cites its review of the immature body of research surrounding the relationship between elevated CO2 and cognitive ability. Half of the studies reviewed indicated that human cognitive performance declines with increasing CO2 concentrations. The problem entailed in this study, similar to the Zanobetti study above in Example 1, is that it does not develop any underlying mechanism which could explain instances how elevated CO2 directly impacts cognitive performance. This is not a condition of ‘lacking mechanism’ (as sometimes the reality is that one cannot assemble such), rather one in which the current mechanism paradigm falsifies the idea. The study should be titled ‘Groundbreaking new understanding on the toxicity of carbon dioxide’. This is of earth-shattering import. There is a lot of science which needs to be modified if this study proved correct at face value. The sad reality is that the study does not leverage prior art in the least. As an experienced diver, I know that oxygen displacement on the order of 4 percentage points is where the first slight effects of cognitive performance come into play. Typical CO2 concentrations in today’s atmosphere are in the range of 400 ppm – not even in the relevant range for an oxygen displacement argument. However, I would be willing to accept this study in sciebam, were they to offer another mechanism of direct effect; such as ‘slight elevations in CO2 and climate temperature serve to toxify the blood’, for example. But no such mechanism exists – in other words, CO2 is only a toxicant as it becomes an asphyxiant.5 This study bears explanitude, it allows for an existing paradigm to easily blanket-explain an observation which might have otherwise indicated a mechanism of risk – such as score declines being attributable to increases in encephalitis, not CO2. It violates the first rule of ethical skepticism, If I was wrong, would I even know it? The authors failed in their duty to inform about the known mechanisms of CO2 interaction inside the body, and as well failed to address standing knowledge about industrial farming nutrient depletion. As well, this study was a play for political sympathy and club rank. Couching this pseudo-science with the titular word ‘Possible’ is not excuse to pass this off as science. Grade = D (inexpert find, harmful science).
Example C – Orphan Question, Failing in All Seven Elements of Hypothesis, and Especially Duty to Inform of Intelligence
A Population-Based Study of Measles, Mumps, and Rubella Vaccination and Autism. Madsen, Hviid, et. al. (Link)
This is the notorious ‘Danish Study’ of the relationship between the MMR vaccination and observed rates of autism psychiatric confirmed diagnoses inside the Danish Psychiatric Central Register. These are confirmed diagnoses of autism spectrum disorders (Autism, ADD/PDD and Asperger’s) over a nine year tracking period (see Methodology and Table 2). In Denmark, children are referred to specialists in child psychiatry by general practitioners, schools, and psychologists if autism is suspected. Only specialists in child psychiatry diagnose autism and assign a diagnostic code, and all diagnoses are recorded in the Danish Psychiatric Central Register. The fatal flaw in this study resided in its data domain analyzed and the resulting study design. 77% of autism cases are not typically diagnosed until past 4.5 years of age. Based upon a chi-squared cumulative distribution fit at each individual μ below from the CDC, and 1.2 years degree of freedom, and 12 months of Danish bureaucratic bias = .10 + .08 + .05 = 0.23 chance of detection by CDC statistical practices – or 77% chance of a false negative (miss). The preponderance of diagnoses in the ADD/PDD and Asperger’s sets serves to weight the average age of diagnosis well past the average age of the subjects in this nine year study – tracking patients from birth (average age = 4.5 years at study end). See graphic to the right, which depicts the Gompertzian age-arrival distribution function embedded inside this study’s population; an arrival distribution which Madsen and Hviid should have accounted for – but did not. This is a key warning flag of exclusion bias. From the CDC data on this topic, the mean age of diagnosis for ASD spectrum disorders in the United States, where particular focus has tightened this age data in recent years:6
• Autistic disorder: 3 years, 10 months
• ASD/pervasive developmental disorder (PDD): 4 years, 8 months
• Asperger disorder: 5 years, 7 months
Note: A study released 8 Dec 2018 showed a similar effect through data manipulation-exclusion techniques in the 2004 paper by DeStefano et al.; Age at first measles-mumps-rubella vaccination in children with autism and school-matched control subjects: a population-based study in metropolitan Atlanta. Pediatrics 2004;113:259-266.7
Neither did the study occur in a society which has observed a severe uptick in autism, nor during a timeframe which has been most closely associated with autism diagnoses, (2005+).8 Of additional note is the fact that school professionals refer non-profound autism diagnosis cases to the specialists in child psychiatry, effectively ensuring that all such diagnoses occurred after age 5, by practice alone. Exacerbating this is the fact that a bureaucratic infrastructure will be even more slow in/fatal in posting diagnoses to a centralized system of this type. These two factors alone will serve to force large absences in the data, which mimic confirmatory negatives. The worse the data collection is, the better the study results. A fallacy called utile absentia. The study even shows the consequent effect inversion (vaccines prevent autism), incumbent with utile absentia. In addition, the overt focus on the highly precise aspects of the study, and away from its risk exposures and other low-confidence aspects and assumptions, is a fallacy called idem existimatis. I will measure the depth of the water into which you are cliff diving, to the very millimeter – but measure the cliff you are diving off of, to the nearest 100 feet. The diver’s survival is now an established fact of science by the precision of the water depth measure alone.
In other words this study did not examine the relevant domain of data acceptable to underpin the hypothesis which it purported to support. Forget mechanism and parsimony to prior art – as those waved bye-bye to this study a long time ago. Its conclusions were granted immunity and immediate acclaim because they fit an a priori social narrative held by their sponsors. It even opened with a preamble citing that it was a study to counter a very disliked study on the part of its authors. Starting out a process purported to be of science, by being infuriated about someone else’s study results is not science, not skepticism, not ethical.
Accordingly, this study missed 80% of its relevant domain data. It failed in its duty to inform the scientific community of peers. It is almost as if a closed, less-exposed bureaucracy were chosen precisely because of its ability to both present reliable data, and yet at the same time screen out the maximum number of positives possible. Were I a criminal, I could not have selected a more sinister means of study design myself. This was brilliance in action. Grade = F (diabolical study design, poor science).
All of the above studies failed in their duty to inform. They failed in their responsibility to communicate the elements of hypothesis to the outside scientific community. They were sciebam – someone asked a question, poorly framed and without any background research – and by golly they got an answer. They sure got an answer. They were given free pass, because they conformed to political will. But they were all bad science.
It is the duty of the ethical skeptic to be aware of what constitutes true hypothesis, and winnow out those pretenders who vie for a claim to status as science.
The Ethical Skeptic, “The Elements of Hypothesis”; The Ethical Skeptic, WordPress, 4 Mar 2019; Web, https://wp.me/p17q0e-94J