32. Normal Distribution

Normal Distribution

Half the women attending antenatal care in rural Tanzania are, on average, under-weight. Interestingly, however, this percentage is exactly the same in UK antenatal clinics. Does this mean that poor nutrition is not an issue in African villages?

  

Facts never lie. But neither do they always stand up to close scrutiny. It should not surprise you, of course, that 50% of women will be ‘under-weight’, if by this you mean that they are under the average. By definition half of Tanzanian women will be below Tanzanian average, and half above. The same logic applies to the UK. The true fact is of course that by UK standards, far too many rural Tanzanian women are malnourished.

My point is that I could, if I chose, weave the words to support whichever tapestry of beliefs I subscribed to. Vital as facts might be, they are worse than useless without correct interpretation.

(By the way, just to reassure all those who struggle when dealing with numbers, please keep reading. This blog brushes past the prickly subject of stats, but in a diligently thorn-proof way. I am well aware that 100 out of every 15 people in the world have dyscalculia, and for those with painful kidney stones, it’s made worse by also having dyslexia.)

(OK. A serious aside this time: On a scale from Asperger’s to broad, creative thinking, far and away the majority of genius lies towards the end of the scale where numbers and the written word are seen in a different way. ‘The gift of dyslexia’ encompasses some of the most important lights in the history of humanity – da Vinci and Einstein, for instance.)

Anyway, the point is that most of us have a touch of dyscalculia when it comes to interpreting scientific advance. The packaging of the stats-message often matters considerably more than what you find if you can be bothered to take off the wrappings. Present the message well, and the whole world is listening. Because of this, for instance, Mother Earth now spends £20+bn /year on statins. In more than 95% of cases, these statins do no good whatsoever, and in some cases, they do harm. (Notice my clever packaging of the anti-statin message, and please don’t stop taking them if you need them. I was just making a point.)

But it is indeed a hugely important point: At this rate of headstrong and introspective spending, in the 30 years it takes Generation X to pass to the next axis, the world will have spent £600bn on statins – almost all of it ‘wasted’. That would pay for everyone on the planet, every single soul, to have, for instance, safe water and sanitation; or an iPad-mini; or a luxury overnight trip for two on the Manchester Ship canal from Thelwall Viaduct to Weaver Sluice.

How in heaven’s name do we prioritise our collective spending like this? The answer is that we are too ready believe what we are told. When an expensive drug or gadget comes along, we can be tantalised and seduced into believing that we cannot live without it. Or, when a cheap or even cost-neutral bit of sense tries to influence the way we live, we can be too eager to dismiss it as ‘unproven’. (It took 30 years, for instance, from the first evidence that smoking was bad for you, for the message to be routinely adopted by doctors. It took even more than 30 years in the case of the deleterious effects of too much sugar. And, on a personal note, it has taken even more years again for me to assimilate the data that Malbec may not completely wipe out the harm done to one's metabolism by sausages. And I’m still not 101% convinced …)

So here’s the rub: we need information, and we need it to be well-interpreted, in order to legitimise and optimise our use of resources.

We need it, yes, but we don’t often get it, and so our choices can be skewed.

This need for the naked truth is particularly the case when our enthusiasm to do the right thing blinds us to potential negative effects of our actions. Our experience of anecdotal benefit and our anticipation of success, might too easily get in the way of learning proper lessons about what we have been doing. In the last generation or two of Aid from the privileged to the under-privileged nations, these lessons have often had to be learnt more than once. As a result, for instance, every armchair philanthropist now knows that we have to teach people to fish rather than giving them their daily mackerel. (A philosophy yet to bear much fruit in some parts of the Sahara.) The disappearance of overseas aid and good will into black holes is still too common a phenomenon in Africa. The rocks of inefficiency and corruption can shipwreck good-hearted dreams, and the swirling currents of perverse incentives and political expediency can cause many a good scheme to founder.

I am building up to identifying a critical insight about the EMBRACE-Tushikamane project: We need to be careful that we are using the resources to best effect.

Part of the answer is evaluation of impact: Did we save lives? Without doing much harm? However, even such a seemingly simple parameter of measurement is a sophisticated business, needing careful thought. If, for instance, as a response to the problems identified by women’s groups, we implement an improved transport system, linked to training of birth attendants to know which women to send in, then many more women might arrive at the hospital gates …

  

… including those who previously would have died at home on the bloody floor of their hut, or under the grubby lantern of a despairing attendant. The project might therefore, at first, seem to be delivering more death in hospital – but less in the community. Indeed, there may be far less overall, without this being apparent. These are tough issues needing a professional response.

These thoughts on measurement are going to need thoughtful development if EMBRACE-Tushikamane is going to have a long-term sustainable shelf life. I am delighted to report then, that Ammalife has found the wherewithal to support a researcher - Helen Williams - on a three-year PhD, devoted to the EMBRACE-Tushikamane project!! She will be carefully observing, documenting and measuring what happens, and in particular will be looking at how we build bridges across all those organisations trying to help. Suddenly, we have a real live hope that the project will produce transferable lessons for saving lives throughout rural Tanzania.

Meanwhile, let me finish by arguing with myself a little. Let's face it, I am the only one authorised to write on the blog site, and there is therefore the danger of portraying just one view of the topic. Fortunately, in this case I disagree wholeheartedly with myself, if I am trying to say that scientific evaluation is only about stats and normal distributions.

Statistics are vital, but so is narrative and observation of what happens. Stats could tell you that your results fall way outside the normal distribution, and that something unusual is happening. But they will not tell you why, so you might not be able to replicate success nor to rectify what seems like failure. An analogy: your data tells you that none of your flat-packed IKEA furniture seems to stay up long after you reconstruct it.  Number-crunching reveals that not a single item in your new bedroom suite withstood the weight of (respectively), an ironically full cup of coffee; an inquisitive cat; and a weary, end-of-day backside. What these stats don’t explain to you is that hammers don’t work on screws. Observation of what is happening can be as important as measuring the degree to which it happened.

And here is a further complexity: There will be times in science when numbers will be not just inadequate, but utterly beyond inadequate to tell the story. When Copernicus inferred that the Earth revolved around the sun, he did not need to check a representative number of solar systems to ensure that his conclusion reached statistical significance. When Watson and Crick tracked down the structure of DNA, they did not have to repeat the task with other molecules to prove it was not a fluke. At the heart of science is observation, combined with understanding about what you are seeing. Statistics are just a means to that end.

Numbers are important. Vital. But we should use the right tool for the right purpose. Sometimes it is analysis of data, followed by wise inference. Sometimes it is insightful observation and description. Sometimes, even science is inadequate to encapsulate wisdom, and only poignancy can capture the moment. I remember many many times, too many to count, when I have had the privilege to save a woman's life and to deliver a baby into her desperate and exhausted arms, and neither numbers nor words would suffice in describing what just happened.

So let me finish with an example of charitable intervention in Africa which reconciles all of these quandaries, which observes, measures, understands, adjusts - and delivers where it matters.  In the last blog I mentioned Water Aid’s wonderful To Be A Girl campaign, to provide villages with water so that the girls can be freed to live more appropriate lives.

The campaign recently wrapped up, and if you don’t cry when you watch the video, you get your money back:

https://www.youtube.com/watch?v=Op4Lopax3EA&list=UU3JVuo2A_Am7XW5tkFZRi5A

In African development, things can grind to ignominious and poorly understood defeat, under a wrapping of dyscalculia. Or they can work wonderfully. Water Aid is one of the successes, and water distribution is the foundation of future possibility - along with education, food, transport, community health, and eradication of extreme poverty. These are the most basic of human needs, sadly lacking throughout much of Africa. EMBRACE-Tushikamane seeks, in one small part of the planet anyway, to turn under-development into normal distribution.