Please enable JavaScript.

Coggle requires JavaScript to display documents.

TITLE 6 (May 2027): In the pursuit of knowledge, what value is there in…

- - - - Data: within humans, cancer risk rises with body size, about +16% per 10 cm of height (Million Women Study, Green et al. 2011, Lancet Oncology)
      - Paradox: elephants have ~100× more cells than humans, so risk should be far higher; measured cancer mortality is 4.81% vs 11–25% in humans
      - What knowers did next: sequenced elephant DNA; found ~20 copies of the tumour-suppressor gene TP53 vs 1 copy in humans; radiation-induced apoptosis rises with TP53 copy number
      - Parallel: naked mole rats resist cancer via high-molecular-mass hyaluronan (Tian et al. 2013, Nature)
      - VALUE: diagnostic (the hidden assumption of constant per-cell cancer risk was false) and generative (comparative oncology; TP53-based therapies)
      - Link to title: exploring the contradiction was more informative than a thousand confirmations of the within-species trend
  - - - Overall success rate of 78% for open surgery and 83% for Keyhole Percutaneous Nephrolithotomy (PCNL)
      - Paradox: open surgery was better for small stones (93% vs 87%) AND for large stones (73% vs 69%), yet worse overall
      - Cause: surgeons sent harder, larger stones to open surgery, so the aggregate reversed both subgroups
      - VALUE: corrective. Forced stratification, confounder adjustment and randomisation; nothing about the body changed
    - - Intuition: after a heart attack, drugs that suppress extra heartbeats (encainide, flecainide) should prevent sudden death
      - Result: in 1,498 randomised patients total mortality was 7.7% on the drugs vs 3.0% on placebo; the trial was stopped early for harm
      - What knowers did next: regulators stopped accepting surrogate endpoints (fewer arrhythmias) as proof of benefit; mortality trials became the standard
      - VALUE: corrective and protective. The RCT is a machine for producing counterintuitive results that clinical intuition cannot
      - Link to title: here the value came from a method designed to let the counterintuitive appear
  - - - Data: 17,414 British children born in one week of March 1958, followed to age 23
      - Counterintuitive finding: the more older siblings a child had, the less likely they were to develop hay fever; early infection seemed protective
      - What knowers did next: decades of work on immune training, microbiome and the 'old friends' hypothesis; allergy prevalence in farm children
      - VALUE: generative, but with cost: the popular 'dirt is good' message oversimplified the science
      - Link to title: a surprising correlation opened a field; its public translation shows the cost side
  - - - Observation: France had low coronary death rates despite high saturated-fat intake; red wine was proposed as the explanation
      - Consequence: a red-wine and resveratrol research and supplement boom
      - Test: Law and Wald (1999, BMJ) showed the paradox was largely an artefact of time lag between diet change and mortality plus French under-certification of coronary deaths
      - Cost: a leading resveratrol researcher (Das) was found guilty of 145 counts of data fabrication and falsification (UConn, 2012; BMJ 2012)
      - Wider pattern: Ioannidis (2005, PLoS Medicine) argues novel, surprising findings are more likely to be false
      - VALUE: negative. The paradox was explained before it was tested; public dietary knowledge was misled
      - Link to title: exploring is valuable only when the first step is testing whether the paradox exists
- - - - Theory: Marshall (1895) imagined demand rising with price for a staple of the very poor; for a century no convincing real case was found
      - Field experiment: about 1,300 poor households in Hunan (rice) and Gansu (wheat) received price-subsidy vouchers
      - Finding: in Hunan, when the rice price fell, rice consumption FELL as families shifted spending to meat; the demand curve sloped upward for the poor-but-not-destitute
      - VALUE: diagnostic. Forced the split of a price change into substitution and income effects (Slutsky) and showed the law of demand is conditional on budget share and substitutes
      - Mirror of Peto: same function, same kind of value, but found only once and only by field experiment
  - - - Intuition: saving more is prudent for a household
      - Paradox: if every household saves more in a slump, aggregate demand and output fall and total saving may not rise
      - VALUE: corrective of a way of thinking, not a fact: micro intuition applied to the macro level fails (fallacy of composition)
    - - Theory: Braess (1968) proved that adding a road can lengthen everyone's journey when drivers choose routes selfishly
      - Case: Seoul demolished a 5.8 km elevated expressway carrying about 168,000 vehicles a day; the predicted traffic catastrophe did not happen and downtown flows improved
      - VALUE: corrective for policy; also pedagogical, because the result is provable mathematically and then observed
      - Mirror of Simpson: both are errors of aggregation, one in data, one in thought
  - - - Rational-actor prediction: the responder accepts any positive offer, so the proposer offers almost nothing
      - Result in industrial societies: mean offers of 44–48% (Yogyakarta, Los Angeles); low offers are rejected
      - Machiguenga (Peruvian Amazon): mean offer only 26%, modal offer 15%, and low offers were accepted
      - VALUE: diagnostic (homo economicus fails as a description) AND a warning: what is 'counterintuitive' depends on the culture of the knower, as with the Müller-Lyer illusion
      - Link to title: the value of the counterintuitive is relative to the intuition it violates
  - - - Natural experiment: New Jersey raised its minimum wage from $4.25 to $5.05 on 1 April 1992; 410 fast-food restaurants in NJ and eastern Pennsylvania surveyed before and after
      - Counterintuitive result: no fall in employment relative to Pennsylvania, contradicting the standard supply-and-demand prediction
      - Dispute: Neumark and Wascher (2000) used payroll records and found the opposite; 30 years of argument with no controlled experiment possible
      - VALUE: real but stalled. The paradox maps a limit of the assumption (competitive labour markets) rather than settling it
    - - Claim: national income rises but average happiness stays flat
      - Re-test: with better cross-country data the paradox largely disappears; richer countries do report higher life satisfaction
      - VALUE: shows a 'paradox' can be an artefact of thin data, as with the French paradox in medicine