Thursday, 9 July 2020

Mastering Science

Late in my working life, I started a full-time, full-on, job at The Institute. Late in my time at The Institute, I started supervising /mentoring research theses for the MSc in Pharmaceutical Regulatory Affairs. The students are a normal distribution, some earnest, some clever, some both, some neither. Some of them are still in touch and some have gone back to their own lives with whatever they got from the degree.

You be sure that only a tiny fraction of whatever they got from the degree ever appeared on the Syllabus or Learning Outcomes. Almost all the stuff you try to cram into your head to pass the final exams pops out over the next week because of elastic rebound. You never really wanted to store that information and it gets displaced by the next hot girl's telephone number; the location of the car-keys; or how to spell paraphernalia depending on your age and priorities.

When the thesis was finally submitted one of my recent mentees heaved sigh and wrote "Hope I’ll get the good mark for my work"

Which was a bit of a trigger for me, because the thing I hate most about teaching is giving marks out of 100. It seems absurd to reduce the tangled weft of the journey to a single number. Let alone a 68%.  Some of my colleagues are super confident at the reproducibility of their assessments: giving 6.8/10 for a lab report presumably in the belief that it's neither 6.7 nor 6.9/10. The somewhat pompous gist of my reply was: Marks are someone else's opinion of your work and they are relevant. But for me the most important thing is for you to recognise that you've done something original and creative, that you have worked hard and revealed something interesting that nobody really nailed down before. When we started this journey 8 months ago, you knew no statistics and had never scraped data from a PharmRegs website. There are only 6 people in Co Carlow tonight who know or care what a Yates Correction is - it's a very select club. By Christmas you will have forgotten about Chi.Sq. (and Yates Corrections) and you may never need to use that test again, but I hope there will always be a warm glow from having learned something new, that seemed difficult, and that you mastered it.
That's why it's called Master of Science.

Wednesday, 8 July 2020

An honourable profession

Katherine Rundell is a Fifty Pound Fellow [for Eng Lit] at All Souls' College, Oxford. She also writes children's books. At the end of last year, Bloomsbury published another book by her Why you should read children's books, even though you are so old and wise. Well, she would say that, wouldn't she? And although it's bound as a book, looks like a book, it more of a pamphlet at ~6,000 words. But none the worse for that and I'm glad this memoir is out there and indeed glad that one copy is in my left hand now.  It was one of two books which were hand-carried, by a circuitous route, to me from Dau.I the Librarian for my birthday last month. Because of its length you can read it through in a sitting; and travel from a dusty public library in Harare, Zimbabwe to the second star to the right, and straight on 'til morning.  Which as any reader of children's books will know, and love, are Peter Pan's directions to Neverland.

I've mentioned Francis Spufford The Child That Books Built before: his memoir is about learning to read [he started with The Hobbit and it was dead slow for the first many pages] and then becoming a bookie-monster devouring everything that had pages and was within arm's reach. Rundell also mentions him and his book with approval. And she also quotes Martin Amis with less appro "People ask me if I ever thought of writing a children's book. I say, If I had a serious brain injury I might well write a children's book". Which is just the sort of magisterial dismissal you'd expect from a self-regarding pretentious git like Martin Amis.

I was a late reader, but spent my teenage years ploughing through a school library which was heavy on  Penguin Classics  and light on froth. Although I read that too when I could find it. Heck, like Somerset Maugham, I read telephone directories, although that was a bit later. There is little to be gained from reading Ulysses at the age of 16; or Augustine's City of God. Because some life experience is necessary to find any ding-dong resonance there. But I was really glad that I found Mervyn Peake's fantasy Gormenghast and held that close to my heart while everyone else was doing Lord of the Rings. LOTR was the Harry Potter of my day and I read it through, twice, but I'm sorry I was too early to get off on the films instead: because there are chunks where Tolkien disappears through his own ring into tedious obscurity. Do we need Tom Bombadill? We do not.

As an Eng Lit academic  - an expert on John Donne, no less - Rundell's book is leavened with wit and literary references "As a child, I had no illusions that children were sweet: children, I knew from my furious heart, were frequently nasty, brutish and short". But she admits to cutting her readers some slack on the vocabulary side: facade, abundance and renunciation make it to the page but adamantine ends up on the cutting room floor. It's like me using the word "salient" in lectures at The Institute on the assumption that my read-nothing students had built up the same vocabulary as my over-educated self. Cutting young readers some slack is not synonymous for dumbing it down to reading age 9. It means thinking a bit about word-choice and not creating an impenetrable clever-me long-word blizzard. Children are quite able to process-or-skip new words as they are encountered. Between the age of 2 and the age of 12, humans acquire a vocabulary of at least 15,000 words. That's a rate of about 5-a-day. If you only read them, you blow the sound of some words. Like the same Dau.I indignantly denouncing her sister for fengeing [feign] sleep in the back of the car.

Should you read children's books? Absolutely, with or without pictures.

Tuesday, 7 July 2020

Crap detector needed

Anybody there? there? there?? There is no "better" time for inhabiting an echo-chamber. It is certainly an extremely fashionable domicile occupied by all sorts of big name Twitterati, Celebs and Influencers [TCI].  Now whatever your virtues, you really don't get to be a major TCI if you are a shrinking violet with a strong sense of your own humility and fallibility. When it comes to public health, it's not a problem if, say, David Beckham or Justin Bieber (who are rich, fit, personable and good looking) make an announcement about diet, mental health or medical treatment. Because you'll discount their opinions with the realisation that neither is, nor claims to be, super bright.  It is of course mildly annoying if you see Gwyneth Paltrow making millions on her goopy "health" products but you can take comfort that it's only money that the buyers-in are losing . . . not their lives.

It's much more worrying when scientifically qualified people start spouting their thoughts on matters medical when those ideas are based more on scienciness than science. Linus Pauling and Vitamin C is the prime example. Life is too short, and brain too small, for us to evaluate, from first principles, every fact and factoid that comes lollopping over the horizon on the TV news or Twitter. Rather than re-invent the wheel again we turn to some authoritative source to see what they think of the new data.

I am a big fan of John Ioannidis, the iconoclastic number cruncher from Stanford [bloboprev I - II - III]. In April he suggested that the Covid-19 death rate was getting exaggerated because not all the undead covid-positive people were being identified. He lamented that there was no good data on this population prevalence. The thought being the deed, within a few weeks he'd recruited more than 3,000 people in Santa Clara County, California and given them a test. That study found that  the rate of infection in Random County USA was maybe 50x higher than the numbers who were known to the authorities. And that meant that the death rate was 50x lower, indeed about the same as seasonal 'flu. Partly because that was putting it to the man and partly because of confirmation bias in my mind that the lock-down was an over-reaction, I bruited the news abroad on The Blob and elsewhere. Finding out the rate of [sub-clinical] infection is really important to develop scientific evidence based policy . . . but the Santa Clara County study is cutting no mustard for addressing the question. So I hereby, far too late, withdraw my endorsement.

You see, sounding off in a we-print-everything-submitted blog or with more brevity on Twitter is easy. Doing the heavy statistical lifting is not. Before I'd even given Santa Clara my imprimatur Andrew Gelman, a competent statistician from Columbia U., had skewered Ioannidis paper as fatally flawed. “I think they need to apologize because these were avoidable screw-ups. They’re the kind of screw-ups that happen if you want to leap out with an exciting finding and you don’t look too carefully at what you might have done wrong.” What Team Stanford had done was assume that the false positive rate [people who test positive but are in fact covid-free] was 0.5% and took their calculations forward from there. What Gelman noted is that 0.5% is the best guess for false positives but they could be as low as 0.1% or as high as 1.7% because the analysis establishing the rate was carried out on only a few hundred people. This is real worrying because Ioannidis had found a rate of 1.5% among his random street sample . . . they could all be false positives! Maybe not all, probably not all, but we are all at sea about what the actual rate of infection is in Santa Clara County or anywhere else. Check out Stephanie Lee at Buzzfeed for a nice Exec Summary of the statistical niceties, written in normal speak. And while you are there check out Stephanie Lee at Buzzfeed noting that there was probably a conflict of interest in the scientific research because a major funder of the Stanford study was running a lockdown-impacted airline.

They say that Justice delayed is Justice denied but the opposite might be true in science. Too much competition means that scientists are all in a rush to be firstest with the mostest. But rushing results into print in peer-reviewed journals or the biorxiv pre-print server, serves science poorly and may result big red faces when retractions are called for.

Monday, 6 July 2020

Toe bone connected

We have one planet, so let's not mess it up!
Ooops, maybe too late. I R [not really] working from home, because my job in The Institute involves ludicrously long holidays, so that means I can catch up on the book-reading and listen to the wireless between episodes of work. One thing I found in teaching a silly variety of courses over the last nearly 8 years, is that everything is connected. Maybe only in my mind; but I can remember several times when I shared something I learned about environmental water chemistry in my human physiology lectures. If we'd had normal library service then Ramp Hollow, Steven Stoll's wide-ranging and super distressing history of Appalachia [prev] would have been returned months ago even though I never finished it.

One of themes running through the book is reflecting on whether people who had no money were 'poor'. The inhabitants of the hollows, and dead-end valleys of Appalachians were subsistence farmers who grew or hunted or scavenged almost everything they wanted or needed from their immediate environment. That included a patch of corn, squash and potatoes, a few hogs and whatever they could find in the woods. That included wood for fuel, furniture, tools, toys and building. For the last 100 years, The Man felt that, for their own good and the good of the country, these people would be $$$ better off working in mines, foundries and factories for money and buying their food in shops like folks in Washington DC. It was the devil's bargain for those who embraced it because the work was relentless to pay the rent, taxes and still have enough for food. It's what has been meted out on peasants the world over as they wash-up crammed into favelas, townships and barrios working for $2-a-day making shoes and shirtings for you and me.

A key plank in the legislation was Wickard v. Filburn, a 1942 SCOTUS ruling that gave an enormous boost to the regulatory authority of The Feds. It hinged on what Roscoe Filburn a small farmer from Ohio was entitled to do with the wheat he grew on the acres he owned. Agricultural Adjustment Act of 1938 limited the amount of wheat that could be harvested by any farmer in order to keep the price artificially high . . . to benefit all farmers. Filburn was entitled to harvest 11 acres which he did; but he also sowed and harvested an additional 12 acres of wheat "for his own use". He argued that this additional wheat never reached the market and so had no impact thereon. SCOTUS ruled, unanimously, that by eating his own bread, milled from his own wheat he was, of course, impacting the economy because he wasn't buying bread in the store in town . . . and that the ripples of this would extend across state lines [key issue in the perennially contentious balance between states-rights and central government] to affect the price of wheat in Kentucky and [{Quiz time betcha cannot name the other four OH-bordering states}]. Since then you cannot really claim to be self-sufficient. Poor old Roscoe, he must have felt like the butterfly that flapped its wings in Costa Rica and caused Storm Ophelia here 3 months later.

That paragraph in Ramp Hollow made me particularly receptive when during my next tea-break I caught Pat Kenny interviewing Garret Kelly, the gaffer at SEE Change Net based in Sarajevo. SEE Change Net seems to be a climate/carbon mitigation expert group with a 2019 turnover of 450,000 Bosnian Marks =~ €225,000. It's hard to make out what they actually do [except buy Carbon Damage Mitigation Certificates (CMDC) when they fly; serve no meat [vegowraps for all R] at their functions and promise to walk a lot] but Kelly seems to know his carbon onions and had been asked to comment on the programme-for-government negotiated among the Greens and Fine Gael and Fianna Fáil.  He reckons that for a small island nation with a really windy sea-board, off-shore wind is really the only sensible solution. We import a climate-wilting €3 billion worth of fuel each year because we are wedded to cars, buses and trucks. He pointed out that our 7 million hectares of land is limited and alternate uses must be traded off against each other. Those uses are broadly:
  • Biodiversity
  • Food
  • Fuel
  • Carbon sequestration in forest
That puts the kibosh on truly sinful ideas like bio-diesel. WTF do you think you're at if you could consider lurrying nitrates onto a field to increase the yield of oil-seed rape to drive some yob's land-rover. Nitrates will ineradicably alter the microbial flora of the field. If you add pesticides to mitigate damage by slugs, flea-beetles and leaf-miners then you are depleting the really pretty bio-diversity as well. Don't, just DON'T. And, me, I should really look at my smug self-sufficient wood-lot as a means for heating the kettle for my hot-water bottle. In the post interview comments, a couple of people pleaded for methane digestors to generate electricity to be integrated into all municipal waste-water treatment plants. You'd have to look carefully at the pay-back time of that sort of investment and probably appreciate that any such contribution is going to be in the ha'penny place in servicing our relentless. implacable, demand for yet more energy to store our holiday snaps and hilarious prat-fall videos; drive to the mall to buy our daily lattes.; and keep the house at 25°C so we can lounge about in tee-shirts watching the rugger in February.
Tsk! and, like, harrrumph!

Sunday, 5 July 2020

Sunday satisfy 5th

Whaaat's happenin'?

Saturday, 4 July 2020

Non-binaire, pronoms

I am The Patriarchy: ticking all the boxes: old ✓, white ✓, straight ✓, male ✓, educated ✓, wealthy ✓, (retired). But I'd like to think I am down with The Gays. A couple of years ago, The Institute dealt with the problem of gender-neutral bathrooms by putting the ⚧ symbol on the wheel-chair access toilets. As a preamble to slaggin' The President about this shameful expediency, I affirmed "I'm down with the gays". This caused considerable flutter from herself and another of my colleagues, who was earwigging "You can't say that" they said in unison. I clarified "I'm down with the gays" not "down on the gays"  and then ripped into the cheap signage solution to what the the management considered a metaphysical issue rather than something that marginalised a chunk of the community.

I've been proactive in my attempts to get down with the Pronouns. And I've added {he | him | his} to my e-mail sig on the grounds that if everybody does it then it's easier for those who might otherwise get othered. Dau.II has been proactive in her Coronarama attempts to get down with a foreign language. As a Foodie, she has naturally concentrated on DuoLingo  French . As a raised anglophone, she's worn out by everything having a M/F gender. [English used to be a highly inflected language like Latin or Irish but we've dumped most of the word endings and all the gendering. The spelinge otoh is bonkers and this makes it quite difficult for foreigners.] "everything" is not just the la/le/les  and un/une/des and the pedantically agreeing noun endings cousin / cousine but the fact that the attached adjectives have to agree as well votre beau cousin but ma belle cousine.

The clash of grammatical and political gender all came up last weekend in a three-way skype with Bob, Dau.II and Dau.I [our Go To  BLT liaison officer and a librarian]. While we gabbed on Dau.I multi-tasked her way to a couple of informative sites about pronoms non-binaire. Seems the most common N-B pronoun is iel which sounds different from il or elle; but it's still fluid, with some folks kiting al ol ul and yul. The ultra-conservative Patriarchal L'Académie française is still reeling from the loss of l'âccênt circônflêxe and has yet to add its imprimatur to iel. Indeed in 2017 the academy had kittens about inclusive language «devant cette aberration inclusive, la langue française se trouve désormais en péril mortel, ce dont notre nation est dès aujourd’hui comptable devant les générations futures» "in the face of this inclusive aberration, the French language finds itself in mortal danger, and today's users will find themselves answerable to succeeding generations".  The Atlantic batters at the gates of Chateau Patriarch.

You can follow up neutrality in several other languages. Big long essay by Amia Srinavasan on English pronouns masquerading as a book [this one] review in LRB.

Gender in French, like speeling in English, is a bit random, as you'd expect from something that jest growed over several hundred years.
Le Boywords
English
La Girlwords
English
blé
Wheat
orge
Barley
soleil
Sun
lune
Moon
sein
Bosom
poitrine
Chest
vase
Vase
vase
Slime
livre
Book
livre
Pound £ lb
fromage
Cheese
viand
Meat
sac à main
Hannnndbag
tie
Cravat
There are rules and conventions but also many exceptions.

Friday, 3 July 2020

Test is cobblers

Philip Nolan is President of Maynooth U and a really smart chap, trained as a doctor, prized as a teacher, acknowledged as a researcher in the neurological control of sleep apnea. As the public spokesman for one of the few Irish Universities, he's often being called up by journalists for a soundbyte on matters of public health policy. And why not? He seems to be doing a good job bringing Maynooth from a sleepy 19thC seminary to take its place as a great liberal arts university. As an active tweeter, he doesn't need to be asked to put something into the public domain. At the end of June, he posted a rather interesting thought experiment on the idea that Fortress Ireland, having sucessfully flattened the curve, could fight off incommmming Covid-Hordes by doing a quick PCR swab test at the Airport which would be no more onerous than inching through Customs, Immigration, Passport Control and Homeland Security at La Guardia.

He was using guesstimated numbers, so the details of his outcome may vary but they are likely to be in the correct ball-park. First off is to disabuse any fond fancy that a quick swab is easily endured. To be done properly [and there is no point otherwise!] the administrator has to insert a loonngg cotton-bud waaay down your gullet, beyond the gag-reflex, and scrrrrape the epithelium. If your eyes aren't watering and your fists balling for a fight response, then it's not down far enough!

No test is 100% accurate, as when Leonard Mlodinow was unexpectedly tested positive for HIV. You'll get both false positives and false negatives as well as getting the answer correct. For a test to be useful the correct answers have to far outnumber the incorrect ones.These two wrong answers can be reduced to two key statistics: the specificity and the sensitivity respectively. Nolan Estimates:
  • 1 in a 1,000 random people is actually harbouring active virus
  • The sensitivity is 85%: of the people who really have SARS-CoV2 the test will only catch 85% of them. There are lots of tests out there: their sensitivity ranges from 70% to 95% so 85% is mid-range
  • The specificity is 98%: of the people who don't have the virus 2% of them will test positive
  • And regardless of the effectiveness of any test, recently [days 1-4] infected people will be "latent": incubating the virus but at levels so low that they can dodge the invading cotton-bud 
  • For the next 10 days the virus will be running its course and potentially detectable
  • Thereafter the human hosts will either be out and about having survived the ordeal, and no longer shedding virus OR fighting for their lives in an ICU bed somewhere OR dead.
  • These data are helpfully summarised into this quantitative flow-chart
Testing, say, 50,000 people at the airport at the airport is going to cost someone something North of €100 each. Nolan calls it £6,000,000. Numbers are down a bit but in 2019, 30 million people arrived at Dublin Airport - about 100,000 every day. On those numbers the annual cost of Covid testing would be €3 billion! It only costs €60 billion to run the whole country. That's going to be a very hard pill for the Finance Minister to swallow and I very much doubt if tourists and executives are going to pay €100 each for the anti-fun privilege of having their throats reamed out. But the other problem identified by Prof Nolan is that the people who are flagged as positive who are in fact whistle-clean far outnumber the actual Perps. 

Numbers on a screen mean so little. For most of us, most of the time, a billion is about twice as big as a million: so Jeff Besos's net worth and the GDP of Hungary are about the same at €130 billion. With that money he could buy 1,000 flash new yachts with a helipad.  So I reconfigured the blocks in the diagram above to a scale of 5 people to 0.25 sq.mm.
To isolate those 30 real epidemiological hostages to Ireland's economic fortune you've got to accommodate 1,000 other totally clear, clean-living folk in whatever pod the new Minister of Health deems appropriate. And remember, every day two such batches of incommmers are going to be tested and kept away from the rest of us. For once, I don't know what the answer is but testing for live virus Ain't gonna happen!
Explanatory note:
Testis cobblers
cobbler's awls
rhyming slang: balls