I couldn’t find a word in between impertinent and impatient
that carried a joint meaning. We dwell in a society of, as Walt Whitman called it, "misshapen beauteous forms." The land is awash with the dutiful bowing to the heavy velvet and the
coruscating diamonds that blink and blind. And yet the wizard within the velvet
has no idea of what is going on, save the faithful that keep bringing gifts with fixed smiles of servitude.
Science is in the throes of such a cannibalistic
transformation. Don’t get me wrong, there are a few innovators out there
tinkering away at the next achievement, but majority of the sad class of
yesteryear seems to have a new-found love for “manipulation.”
Ah, the class of yesteryear continues to shine in the field
of bastardizing scientific rigor, you know, the kind that helps make their kind
replete with financial progress. These minted “old know-little but act-big somethings”
have perfected the art of tom-foolery. Why toil for months and months or years
on ponderous and excruciatingly difficult experimentations, verifications and
ultimate validation only to have a 50-50 chance of “success” and a “bust” lurking
on the horizon, staring back at you in the face with the certain feel of time,
money and effort loss. Would one go the way, knowingly? Especially given the
“get rich quick,” “trade like a millionaire” books flying off the shelf. And if
you haven’t noticed Bitcoin at $15,000.00 per imaginary coin life we are living today.
There must be a better method? One asks.
And voila, the answer is there!
The phoenix of little work and large gains has arisen. The
hard work is a thing of the past. Now it is a matter of a few clicks, a few
number “manipulations” and the results are monumentally deafening in their
emergent applause.
Today, scientific discovery is based on computing power and
a few biased and sometime shady statistical methodologies. Flawed and
pseudoscientific misleading research is ruinous to societies and create unwise
regulations. Usually a “strawman” (an intentionally misrepresented proposition that is set up because it is easier to defeat than an opponent's real argument) hypothesis is advanced, for example, two treatments
A and B are equally effective, the two treatments are then compared, and any
p-value less than 0.05 (p<.05) is, by convention, usually considered
“statistically significant” and that tends to disprove the strawman hypothesis and
prove that effects of the treatments are similar.
When scientists explore many questions to look for their “shiny”
needle in a haystack (That they will find (false positives) through
tortuous manipulation of the data) that is known as “data dredging.” (Data dredging - also data fishing, data snooping, and p-hacking is the use of data mining to uncover patterns in data that can be presented as statistically significant, without first devising a specific hypothesis as to the underlying causality).
It is wise to ignore studies that pose too many questions.
Another mechanism utilized commonly is the desire to
formulate a hypothesis after the analysis is done, which is known as HARKing- Hypothesis After the Result is Known. https://www.ncbi.nlm.nih.gov/pubmed/15647155
These mechanisms create false-positive-generating machines.
Remembering that when 17 comparisons are made there is a 5%
chance of a false positive leading to a 58% chance of the false conclusion that
at least 1 difference exists!
Subverting the self-correcting nature of the critical
inquiry in science can and lead to a decade-long politically-sustained funding
source for this pseudo-scientific scholarship. This exercise is furthered by
the consensus-seeking crowd who offer mostly rhetoric in response to critical
inquiry and ultimately hurl ad hominems when pushed to the wall. This collaborative consensus-based censorship
continues to keep the revenue stream from drying up.
As Marcia McNutt, President, National Academy of Sciences
stated, “Memory-based research methods produce self-reported data that are
implausible and should not be used to establish the Dietary Guidelines for
Americans or regulate the $5 trillion US food industry.” A reminder for those watching television and ogling at magazine ads; chocolate does not cure cancer and sugar only tastes sweet but has some nasty properties against the health of the body.
Moving through the statistical landscape of selection,
scientist-based, industry-based and political-gain-based bias, we arrive at a
slightly better form of rigor, but no less flawed…
Let us for instance take into account the “Confidence Level”
to start with. I have written about the p-Value
HERE. But let me take you on a different excursion.
Assume you are an individual within a neighborhood, within a
city, within a county, within a state and within a country and within the
world. Ok you don’t have to assume, you are! You have been living there for
quite a while. Hence as Lamarckian twists would have it, you do share the
plentitudes of genetic markups that have happened in over that period of time.
Assume you arrived from another country and brought a package of genes from
that country and your progeny spread in different states. Does that mean that
the host nation now shares the same genomic makeup within their genetic and
other epi-genomic structures, with you? Any half-wit will tell you, No! You
have your own genomics cobbled together by your parents and that from your previous
and present environment and your life style.
So, when we use the Bell curve as a standard and quickly
create the means and standard deviations to ascertain the z-values
Are we not creating a virtual genetically homogenized
society that belongs under the same Bell Curve? And the Mean or center of that
Bell curve then constitutes the entirety of the world population and
furthermore if the statisticians and clever physicians believe that within that
95% CI, the chance of not conforming within is essentially considered an outlier status (a
rare event). Well, if you follow the argument, the case for the world
population mean in genomics and infectious agents has never been tallied and
with 7.2 Billion it is unlikely to ever be accurately tallied to discern the
real mean and therefore for comparative purposes ever to be the central part of
the Theorem to give you that lusted after 95% CI, the confines of which entitle
one to say that the Ho (Null Hypothesis) is rejected. (In other words the outlier status might encompass a much larger population then is deemed accurate in the 95% CI of the two statistical deviations. So, to say that we have
95% certainty that Treatment A will work in a large subset of the population is patently false and falls flat. Because the genomic/epigenetic structure of the group does not guarantee enough variability to have an
assigned mean. If there is no real assigned mean then sampling error goes out of
the window as well as the standard deviations as well as the mean of means.
Don’t they? The reference point is therefore a false premise from the outset!
That felt like gobbledygook, didn’t it, perhaps? Let me try another
way… It is different to take a Rajeev from India, a Tolstoy from Russia, a
Smith from England and a Lee from China and consider their genomics (exposure
to viral inclusions and other Lamarckian forces) as similar enough to
constitute the porridge of the central theorem with its 2 standard deviations.
Since we have determined that if the real mean is not known, how then can the
STDEVS be? Now if we cannot do that, then the results of the Null Hypothesis
generating studies flies in the face of reality with speed. The n can go to a
1000 in a society (a large number for a study) even within the melted-pot social
variety and still not have real homogeneity to give us a good calculated
probability finding. No wonder landmark studies (47 in 53 or 89% in AMGEN’s study)
done on cancer patients in different countries bear different results. From
academe to the real physician work, the translation of a drug activity and
benefit fails to consistently show similar advantages. All one has to do is
look at the Immunogenic forces on the Hepatitis B virus in different parts of
the world that have yielded different Hepatitis B genetic subgroups of those
viruses (Baltic population with Hepatitis B "D2" variety and the SouthEastern with a Subgroup "C" and in Brazil with mostly an "A" subgroup variety. And the subgroup numbers keep getting longer and longer with time and
exposure. Diversity is nothing more than a mutation. And a mutation means a
different host against a different guest. And for any species to survive and thrive it needs to diversify through natural selection, not by artificial means.
So how come these impertinent imposters that play with our
reality under the guise of science and write papers upon papers (mostly to get
tenure and prestige) don’t tell the truth. Perhaps the revenue stream of the
sponsor is a gift that keeps giving?
Am I biased? Sure! Aren’t you?
To pass on our biases with a certain degree of gravitas is
what makes this whole game kind of sinister. It would behoove the “experts” to
stop rattling the chains of “Look at Me, Look at Me, I have a new paper,” and
do some real unadulterated science for a change, that is, if they still know
what real science is.
Take this with a grain of salt, too. Remember, I have my own
personal biases.
No comments:
Post a Comment