Why Are Statisticians Silent?

I regard Epidemiology as not science – it is counting. And yet I regard Statistics as science, even though it is counting. What is the difference?

I think that the difference revolves around interpretation. Statisticians count incidences but do not derive ’cause and effect’. Epidemiologists also count incidences but derive from those incidences ’cause and effect’. Thus, in simple terms, if more smoker get LC than non-smokers, then smoking must have caused the LC – according to Epidemiologists. But a Statistician would derive no such conclusion. He would ask what proportion of smokers died from LC as a percentage of ALL smokers, and what percentage of ALL non-smokers died from LC as a percentage of ALL non-smokers. But he would still not presume ’cause and effect’. He would enquire further and possible compare differences in rural areas with city areas, or windy cities on the coast compared with stagnant interior cities. The complexities are numerous.

But we hear nothing at all from ‘The Statistical Society’, or whatever it might be called, about the abominable ‘overweight or obese’ blather.

One can only conclude that the Statistical Society is either colluding or frightened. In either case, that is very sad. In fact, it is cowardice of the highest order. As a proper ‘science’, it should be fighting as hard as it can to retain its credibility. In order to do so, it must rely upon voluntary donations rather than Government grants. All organisations which rely upon Government grants are at risk of interference. Perhaps that was the reason for ‘Charities’ receiving grants in the first place – they can be controlled. Perhaps that is the reason that Gov will not conduct a ‘bonfire of the quangos’.

Finally, it is perfectly obvious that the current ‘political parties’ are ‘not fit for purpose’. Neither of them really advocate freedom. They advocate freedom ‘within the corrupt and totalitarian, one size fits all, law’.

Statisticians could throw a spanner in the works if they had the courage. What exactly is the incidence of LC in spouses of smokers as compared with non-smokers? The importance is statistical in the sense that ‘no difference’ cannot be denied.

Zero means zero. Only if evidence could be produced that the ‘zero’ is wrong, could ‘danger’ be presumed.

In my ignorant way, I have tried to show how TobCON has diverted, corrupted, and abused ‘science’, and that those who could have stood in opposition, true scientists, have chickened out.

There is so much political junk that only the destruction of the major parties in the UK, and re-formation of them as distinct entities, can unite ‘The People’.  It is too easy to demand ‘freedom’ in a vague sense. It will not work. It never has. ‘Freedom’ means that entrepreneurs (businessmen) can offer smoking pubs. It is important to understand that ‘SHS danger’ would be accepted in those circumstances.

So what has ‘The Statistical Society’ got to say about SHS?

Maybe tomorrow I shall enquire.

Advertisements

11 Responses to “Why Are Statisticians Silent?”

  1. garyk30 Says:

    Stats folks will present all sides of a study; not, just the side that seems to prove their argument.

    For instance:
    Smokers may have a higher lung cancer incidence rate; but, a stats person would also include the data that the data shows that smokers have the same probability of not getting lung cancer.

    Say comparative rates of 1/10,000 and 10/10,000.
    Not getting rates are 9,999/10,000 v 9,990/10,000.

    9,999 is only 1.0009 times greater and a true stats person would note that as being the same.

    Only people with agendas will ignore the paradox in data here.

  2. Smoking Lamp Says:

    It would be really cool to get the source data for gary30’s statistical paradox. It seems to me that that data is a good start for discrediting the tobacco control lies.

  3. garyk30 Says:

    Heavy smokers and never-smokers have almost precisely the same chances of NOT dying from those diseases caused by smoking.
    Doll’s doctor mortality report.
    http://www.bmj.com/highwire/filestream/400720/field_highwire_article_pdf/0/bmj.38142.554479.AE

    The table on page 3 shows this:
    Lung cancer deaths per year.
    heavy smokers(25+/day) = 4.17/1,000 = 995.83 did not die.

    never-smokers = 0.17/1,000 = 999.83 did not die.

    999.83 divided by 995.83 = 1.004.

    Never-smokers are only 1.004 times more likely than heavy smokers, to not die from lung cancer!!!

    When you have to go to 3 decimal places to find a difference, that difference is, for all practicality, non-existent.

    Other results:
    mouth/throat cancers = 1.001 times more likely to not die.

    all other cancers = 1.002 times.

    COPD = 1.002 times.

    other respiratory = 1.002 times.

    heart attack = 1.005 times.

    stroke = 1.002 times.

    other vascular = 1.003 times.

  4. Smoking Lamp Says:

    Thanks Gary30! Good stuff!

  5. Some French bloke Says:

    From the other perpective, one that is no doubt favoured by antismokers, 4.17 (heavy smokers’ yearly rate of LC death per 1,000) divided by 0.17 (never-smokers’ rate) = 24,53 (widely publicized as smokers’ relative risk).

    Isn’t it fascinating that from the same set of data Doll et al used to extract their oh-so-scary X 24 relative risk of LC for (heavy) smokers, also come garyK’s quite unimpressive X 1.004 chances for never-smokers of NOT getting LC! This paradox should figure prominently in any contemporary textbook of epidemiology.

  6. junican Says:

    Thanks for the comments.
    SFB. There is no real paradox. If you go swimming, there is a chance that you will drown sooner or later. If you never swim, you will never drown. Thus, the incidence of ‘going swimming’ is critical. None swimmers (and non-smokers) have zero risk, and are therefore bad comparisons. Only differences between smokers matter, in which case things like country and city and other variables become very important.
    People who never fly will never die when an aircraft crashes. What use are they as a comparison as regards air crash deaths?
    Statisticians know full well that TobCON distorts statistics. Why are they silent?

  7. Some French bloke Says:

    There is no real paradox.

    I just borrowed the word “paradox” from Smoking Lamp’s comment, as an apposite term to describe any fact or assertion that goes against the grain of accepted popular ‘wisdom’ – and sadly at the moment, anti-smoking prejudice fits that description.

    Non-swimmers (and non-smokers) have zero risk […] People who never fly will never die when an aircraft crashes.

    Non-smokers have an elusive, ever-fluctuating baseline risk, as opposed to non-swimmers’ and non-flyers’ absolute zero risk. Light, moderate, and heavy smokers also have a maddeningly variable relative risk according to the U.S. and U.K. antismokers’ favourite studies.

    Couldn’t the fact that urban non-smokers are, willy-nilly, ‘swimming’ in a pool of cancer-causing chemical compounds, just like their smoking counterparts, start to explain the existence of that baseline risk?

    Remember that, according to GaryK’s analyses of the (in)famous Doctor’s Study, 84% of non-smokers get all of the so-called smoking-related diseases and conditions 85% of smokers suffer from, only in different proportions and time frames.

    I am reminded that Percivall Pott’s young chimneysweeps in the late 18th century later developped scrotum cancer in their 30s or late 20s, yet not all of them did, so no absolute risk involved there, since not ALL chimneysweeps got it, though the relative risk was in the hundreds (beyond the antismokers’ wildest dreams). That is real relative risk.

  8. Frank Davis Says:

    Royal Statistical Society

    http://www.rss.org.uk/

    I got interested in them a while back:

    https://cfrankdavis.wordpress.com/2015/10/28/a-statistical-investigation/

    • junican Says:

      It strikes me that the Statistical Society has evaded politicisation.I vaguely remember an attempt by Blair to interfere with Nat Stats, but the boss refused. Nat Stats exists to present FACTS, and nothing but FACTS. As far as I know, it is still pure.
      Perhaps that is the reason that statisticians are keeping their heads down – they do not want to be involved in politics at all, either positively or negatively.
      What we are currently experiencing is ‘fake statistics’ to support TobCON, but REAL statisticians do not dare to object.

      I have said before that it is likely that retired or about-to-retire people who will eventually blow the whistle. The more that TobCON moves beyond facts and truth, the more certain that they will, eventually, be defunded and destroyed.
      It will be up to pubs to lobby for smoking rooms, and they will eventually succeed.

  9. Philip Neal Says:

    The RSS gave Doll a tough time in 1970 and Burch a sympathetic hearing in 1978. I suspect that many statistically-trained people know how patchy the case against smoking is, but long ago decided there is no point defying conventional wisdom.

    • junican Says:

      I agree. But perhaps many of them agreed that smoking was harmful, despite the crap stats created by Doll et al. No motivation to speak out therefore.

Comments are closed.


%d bloggers like this: