Pages

Sunday, December 23, 2018

Hong Kong


Cadaveric workshop Prince of Wales Hospital, Sha Tin


Temple St Night Market, Kowloon



Sunday, December 9, 2018

Curious but not that bright (on measurement 3)


“You are stupid,” he said. “You are so stupid.”


Not exactly what you’d expect when acknowledging a fellow human being during an early morning run. Admittedly the chap was looking a little sozzled and had something or other stuffed into a brown paper bag but it still took me by surprise. I smiled, put my head down, and continued up the hill with calls of “stupid” chasing behind me. My regular run is an out and back loop so after a bit of quiet time it took me back pass the same chap on the return leg.


“Smoke some weed man. Running is bad for you,” he shouted.


“Don’t be stupid man. Smoke some weed!”


Well, I’ll be damned. 


The man has a point.


Data-driven. Now that’s a word combination that troubles me. I want to get to that but first a belated preamble to this series of monologues.


I recently bought a new road bike and it has proven to be an indulgent piece of engineering for an average poddler (even if the gear changers need a recharge every now and then). But it is a proper race bike and comes fitted with an ANT+ sensor that allows data collection of speed and cadence and has the capacity to deliver this wirelessly to any ANT+ compatible display. My numbers are certainly nothing to shout about but I thought maybe it was time to see what they are now compared to what I remember them to be “back in the day”. The cheapest display unit I could find was $220. Well.. bugger that. My curiosity lies well below that asking price.


This got me thinking about why  we go about measuring things. Since the kilogram standard referenced from a block of platinum was made redundant three weeks ago I’ve been thinking about it a bit more. 


https://www.nationalgeographic.com/science/2018/11/kilogram-forever-changed-why-mass-matters/


A child might ask why we measure certain things but not others. Why don’t we measure a hug or a kiss? Surely they are more important than the price of potatoes, or the return on investment, or power output, or lap times. But are they? Logically speaking such value loading depends on whether you need to worry about each or any of those four quantities. A wealthy, practicing surgeon probably has less concern about the financial returns of his hobby farm than a recently retired council worker looking at the return on his annuities. A career cyclist probably takes a keener interest in fluctuations of his power output than a career accountant. A single mother of four probably scrutinises the price of potatoes more so than an executive couple. Or, at least, one would think, that is how it would be. 


To answer the child I would say that there are many valuable things that we simply cannot measure - at least not with the technology available today. But an adult having the perception that he/ she is on the right side of two easily measurable quantities - time (a physical constraint*), and money (a societal construct) - certainly helps with the return on investment of that child’s hugs and kisses. But that’s a topic for another day.


Humans, being curious, analytical creatures try to measure everything. With the increasing availability of cheap and accessible data storage it also means we document a lot of what we measure. The question I have is whether the act of measurement and documentation subtly - and sometimes not so subtly - changes the dynamics of the human experience. A photograph is a form of data collection and storage that contains a lot of information. It is also a measure (or, more accurately, holds numerous measures) of a moment in time. When I take a photo on a bicycle ride or a camping trip I feel that this simple act of documentation changes the experience I have with that environment. I’m not talking about a formal photoshoot or an Instagram fame grab, just the run of the mill snapshot. I’m sure I’m not the only one who feels this.


Many activities are centred around the act of measurement. Scientist and researchers do this day in, day out, as do horologists, historians, surveyors, stock brokers, professional photographers and plumbers. Civilisation cannot exist without measurement, documentation and data. We measure, we document, and we store information of all types. It is the foundation of mathematics, science, engineering, and economics. It establishes a rules-based order in societies, sports, financial services and many other institutions we value. In essence, we can’t live without it. 


Just thirteen days ago the InSight lander touched down on the surface of Mars. Nothing seems to capture the capacity for human endeavour quite like the pointy end of science. And none more so than the physics of space exploration where the metrics are more or less knowable.


https://mars.nasa.gov/insight/timeline/landing/summary/


But measurement, documentation and data also makes an impact at a more personal level. After all, scientists and other folk who immerse themselves in quantitates still have to go home after work. 


Measurement allows one individual to relate to any other individual. A 5 kilogram weight means something different to a weaker, smaller person than a larger and stronger one but it serves as a reference point that both can agree upon. I suspect that 5 minutes also means something different to a person whose internal clock runs quicker than to one whose internal clock runs slower - ie 5 minutes might represent different opportunities to a hyperactive individual than someone who is chilled out and relaxed about life. But time schedules mean we get to meet at the same place at an allotted time. Defining things in terms of their quantifiable measurements allows us to understand our surroundings and communicate in terms we agree upon. But it should not diminish the qualities that are not measured.


And that’s my point. Only a fool or inchoate wants to dismantle the foundation of reason and rules. Yet it is worth thinking about why and what we choose to measure. Measurement is an analytical tool. It (almost always) never fully describes the event or object being measured. 


Take biometrics. This is an area of active consumerism that has exploded over the past few years. Whether you are trying to lose weight or achieve a PB (personal best) biometrics can help gauge your state of progress. But what exactly are we measuring? Take one of the oldest biometrics - BMI (body mass index = weight in kilograms/height in metres²) - as an example. BMI is used as a proxy for health and, by popular extrapolation, happiness. Almost everyone (intellectually, if not viscerally) recognises that weight is a blunt tool for everything other than what it measures, ie body weight. Combine it with a measure of height and it gains greater relevance. A BMI of less than 18.5 or greater than 24.9 suggests that the individual could probably be more healthy. But if you fall within the healthy range of 18.5 - 24.9 then your BMI doesn’t really tell you much at all. As to whether you are happy or not it tells even less. A BMI in the healthy range also doesn’t negate the detrimental effects of smoking, cancer, or participation in collision sports. What about resting heart rate? Blood pressure? Vertical jump distance? Functional threshold power? Lactate threshold? Tidal volumes? Vital capacity? VO2 max? Cardiac ejection fraction? Calcium score? These biometrics have relevance to extrapolations on health and fitness. But these measures are all one dimensional - it takes a degree in the health sciences to put the information together. To be honest most of us haven’t the time or energy to go to such depth. 


Inquisitive? Yes. Curious? Well.. generally speaking, yes. Just not knowledgeable or bright enough to make sense of all that data. 


But not stupid. Nevertheless I think I get the sentiment that the sozzled chap was trying to convey in his sozzled sort of way. People in my suburb with focus and drive - ie a people who run when they don’t really have to - have a disconnect with the vibes that matter. They live in a constructed world that has lost touch with the human condition. A concrete arcade where solace is materialism and escape comes in the form of mindfulness and meditation. I’m not a big drinker and I’m certainly not about to start smoking anything. I run because my personality is not predisposed to a meditative state. It’s always dark outside when I lace up my shoes. In winter my breath smokes the air under street lamps. The cold cuts through my clothes but I know it will warm up when I get going. In summer it’s too hot to run after the sun rises. And I need time to cool down before going to work. So I start in the dark. In any case, all I have for company are my thoughts, my legs, and the pounding heart inside my chest. I run because it is good for me. I like the quiet solitude and the sound the wind makes when it passes over my ears. Ergo, I don’t collect biometric data. The sozzled chap, of course, doesn’t know this, but I’m already on board with his weltanschauung. Sure, if the aim is to run faster or jump higher then there are metrics to take into account to help achieve this. As we get stronger and faster the metrics all start heading in the right direction. Even if we take one or two metrics and work on improving them then this drags up other metrics that might not be measured or even considered. But if we aren’t actually racing then what’s the point of all this measurement and activity?


Self improvement. 


Hmmm. Take a moment to think about what this means and where this leads us. Let’s look at a couple of examples. Say an overweight person decides one day to lose some weight. Another person is comfortable with her weight but wants to run faster over 5 kilometres. Both see endpoints beyond their current state. In marketable terms they have set “goals”. Say the overweight person has a BMI of 25 but is otherwise in strikingly good health and the runner places somewhere in the middle of Park Run (a timed 5km fun run) but otherwise doesn’t race competitively. Both monitor their metrics (weight and run times respectively) in order to better them. They start out enthusiastically and make gains after the first few weeks. They feel great and every interaction they have is seen in a positive light. Life is good. Then something happens. The person trying to lose weight has a few hiccups and starts gaining weight over a week or two. She stops telling other people about her weight and tries out a different diet and then a different exercise regimen. She loses some weight then gains some weight. She looks at herself in the mirror more often than she normally would. She doesn’t like what she sees. She starts buying diet pills over the internet. Her friend, the runner, has managed to get herself regularly placed in the top 30% of her local Park Run. She decides she wants to do better and treats herself a sports watch with an integrated heart rate monitor, GPS, tracking profiles (for different activities) and predictive functions. She gets to a stage where she always places in the top 20% but feels that the exertion to go further is is just a bit too much for where she is with work and social activities. Her watch tells her as much with the stats it has accumulated. She takes up cycling and meets new people. She is in a different place now than where she started. In fact, they both are.


Ignoring the possibility of a deterministic universe or an innate, natural rhythm unique to each and every individual (both of which are relevant but beyond the scope of this monologue) these scenarios - which may play out in all sorts of ways - are easily relatable. Metrics seem to work best when they are processed by the analytical part of the brain. Literally, dragging them out of the fast, instinctive, qualitative part of the brain and into the slow, deliberative, quantitative part seems to optimise measurements to a desired result. The more effort you put into something the more likely you are to engage the analytical part of the brain. Measuring your run time over 5km takes some effort (factoring heart rate, split times, terrain, energy levels, temperature and humidity, etc). Weighing yourself first thing in the morning takes a smaller amount of effort. Feeling fat takes no effort. There is also a key difference between the two girls subtly introduced in the previous paragraph that you might not have noticed. The runner had already committed to running faster before buying the sports watch with all the gizmos attached - nothing unusual given the premium price for a watch with such features. In a sense the the runner was already self-selected to improve her run times before she bought the fancy watch. Data then played a useful but subordinate role in the case of the runner but may well have had a more sinister, subliminal message for her friend. At the non-elite level, the measure of weight loss also encompasses a broader range of relevant, confounding variables than the measure of a faster run time.


It is one thing to understand the value of data and quite another to be obsessed by it. Comfort can certainly be found in measurement and numbers and that’s not necessarily a bad thing - it provides a useful anchor when everything else shifts and sways. But a life focussed on numbers is not healthy and is often unsustainable. It is also not a full life in the broader context of possible experiences.


https://www.youtube.com/watch?v=7_2bybJWQFg


Fitness bands and other wearables that track health and activity markers lie somewhere between the weighing machine and the specialist sports watch. From what I can tell, they have become extremely popular..


I should probably tone it down a bit. Measurement, documentation and data have no moral weight. They can’t be good or bad because they merely serve as a record of what is out there. But context matters. In the right setting sieving through data can be very satisfying. Data can be confirmatory, revelatory, intriguing and entertaining. It makes for good banter - even more so when the information isn’t granular enough to be truly enlightening. A bunch of cycling enthusiasts discussing their latest exploits might be a good example of this. Step it up a notch to, say, professional cyclists looking to win races and the discussion becomes a bit more serious. Up a notch again and we might be looking at things like data from the National Joint Replacement Registry. Further up the chain we eventually get to data on wealth inequality and global warming with more people and more places becoming affected in more consequential ways. 


This started out as a monologue on measurement. And specifically how this impacts  me  us at a human level. I’ve discovered over the course of the past three weeks that you can’t discuss measurement without touching on documentation and data. An increasingly data-driven world troubles a pre-internet relic like me as the measurement of everything that’s measurable adds gravity where it is not required (or desired) and redirects attention from things that are important but immeasurable (or measured imprecisely/ indirectly with current tools). With powerful commercial interests servicing and nurturing a seemingly endless craving for ever greater convenience I can’t see this model for growth slowing down anytime soon. Younger people who have grown up during the maturation of this new paradigm are probably more at ease with its machinations than people like me. Nevertheless they need to be mindful of their place in a data-driven world in order to navigate a meaningful life within and without it.


So, back to the beginning. $220 buys plenty of things. Check out what’s available on an internet-enabled device with a decent search function. One such item is an ANT+ compatible display which, come to think of it, will probably need an upgrade in a year or two. 


It’s not for everyone.




* time in this context is probably more accurately described as a physical dimension often constrained by societal constructs 



Sunday, December 2, 2018

Conjoint Analysis (on measurement 2)


This might be the most interesting article I’ve read all year. Not because of the result it came up with (it doesn’t tell us much that we don’t already know), more for the process by which the research was constructed. It is instructive and yet unsurprising that conjoint analysis is a valuable tool in market research for products and services. 


For a topic as dry as the PIPJ of the finger it is also exceedingly unwieldy. 



















It is interesting to note a few things:

- despite the careful survey design to select the discrete choice experiments (DCEs) utilised in this study nearly 40% of initial respondents did not complete the survey

- of those who completed the DCEs most were white, female, well-educated, and reasonably well off

- even with this demographic the researchers were concerned about fatigue and the employment of heuristic approaches that could confound the results

- algorithms that track internet use combined with artificial intelligence (AI) can garner far more information (deeper in detail, broader in scope, and truer to the users behaviour than what he/ she thinks and says) than any cleverly constructed survey by a human researcher