On Pictures #It'sOn

A picture is worth a thousand words. And herein lies the problem. An image transmits information under the surface, below consciousness. You can be disarmed, misinformed, manipulated if you don't know what you are looking at. Catfished, as the modern expression goes. But I am not talking about Tinder dates. I wouldn't know where to begin. I am a card-carrying boomer when it comes to online love-seeking. I prefer real life. Interestingly, one of my closest friends, or I should probably say was a friend, wrote the premier book on online dating in Britain, Love at First Site. I got to live out vicariously all her escapades as she was finalizing the book, during the editing rounds before publishing. That's when we first met. I was her upstairs neighbor with a newborn to rear. She loved babies, perhaps not me so much. It was instant love, for my son. That was good enough reason for me to call her a friend. Incidentally, speaking of catfish, she gave me a complete lecture on how to present your profile image on dating sites and not lie with your picture or age. Believe it or not, there is an error rate. There are parameters for fudging your stats just so. Just don't call her Hyacinth. Like a middle class debutante keeping up with the Jones, data shown as charts are a minefield of lies. We have truth in advertising but not in data. We lie with numbers all the time. Some times, unintentionally. The most egregious are the professionals wielding statistical analyses like a weapon of mass destruction. Colin Powell, the deadly uranium was not in Niger. It was in the state-of-the-art science labs just down the street. Scientists have been let loose on the unsuspecting public with their endless studies that prove coffee is bad for you one day, and then good for you the next. How can any of this be right? Easy, it is that famous quote, "lies, damn lies, and statistics". What science has forgotten is the integrity of the scientific method, the importance of experimental design. You have to identify the variables, establish the cause and effect theory which is the hypothesis. Then, verify you have articulated the null hypothesis correctly, which is the status quo, giving you your bearings. Without proper controls, how will you know where you are in the grand scheme of things? How will you ensure you found what you claim you found? Once the experiment has been executed, the steps taken, the measurements recorded, the math has to make sense. The finding must not be a confounding of unrelated events, or inadvertent contaminations. Everything must be controlled and contained, for it to sink up with your statistical model. In science you never prove anything true. Statistics is the language of probabilities, confidence levels and acceptable standard deviation. Instead, we say the likelihood of the finding being a fluke of nature or a false positive is so low (i.e., the p-statistic) that we can take the risk of the theory being wrong. Until someone else proves otherwise. Or someone else arrives at the same conclusion we did. An independent checking of the results is a must. Maybe multiple times if the phenomena is controversial. Science not only allows for the possibility of being wrong; it is inherently an error-based system. Now we have studies that are somehow all statistically significant, proving this and that as if it were a new-found religion. People are actually saying things like "believe in the science" and "trust the numbers". Scary times, indeed. The issue is they do not appreciate the scientific process or fully grasp the skepticism required for the scientific mind. Are you aware of the Stanford mathematician Ioannidis who conducted a review of the math employed by a cross-section of prestigious scientific journals only to prove that most of the  math doesn't check out? Which reminds me of something I have been noodling. Can we lobby the ISO 9000 folks to create a standard template to force standardization of scientific papers? That way, labs, whether they are universities and corporations can't play fast and loose with the method and numbers. I even picked out a number for them. It's ISO 9888. Auspicious, right? This is no joke, we need to force a framework for scientific reporting to disclose their thinking. That way we can spot the crime, the underhand maneuvers, the lie. Our medical institutions and food systems are all based on faulty logic. No wonder our health outcomes are down not up despite the innovations (mostly courtesy of the engineers, btw). When was the last time your planes didn't work, your tech didn't compute? Somehow we accept people dropping dead like flies from bad medicine, iatrogenic for the fancy folks. People aren’t dying en masse from bad technology. Heck, most of the equipment used in hospitals and doctor’s office like x-ray machines and MRIs are made by engineers not doctors. I am an engineer by training so I am biased. Why all the bad science? Obviously, it is bad incentives. There  isn't funding for losing ideas. Corporations aren't funding studies to show their drugs kill. Everyone doing science these days isn't truth-seeking. They are all backing into "the answer". Of course, eggs are bad for you. Here eat our cardboard which is cheap to produce but makes us a killing in profit margins. You body is inflamed from the "inhumane" food. No problem, here is a pill that will fix it. Bad side effects? There is another one. And another one. Until you are dead. Oh well. Plenty more humans to milk for money. I once had a heated discussion regarding bad numbers and pretty pictures that did end in tears at work. The culprit. A simple pie chart. I pointed out that the 3D rendering of the chart, although "sexy" and "pleasing to the eye", is ultimately a false representation of the data. The "thickness" or the sense of volume presented in the picture is arbitrary, without backing in hard measurement and therefore inaccurate. My colleague begged to differ. Even though I was the team lead and I made the final call, she refused to comply. Such insubordination, much disrespect. In the end, I had to call my contract manager (not boss) to handle the whole event and literally calm her down like a kindergartener who missed her nap. Then the manager proceeded to give me tips on how to handle personalities like my team member with exact instructions of dos and don'ts. I was bewildered. I would never treat my teammates like how this woman treated me in a professional setting, let alone a supervisor. Don't get me wrong. I have made the hard sell to managers many times in my life and been struck down without so much as a reason why. But true to my "victorian" upbringing of "children should be seen, not heard" and "if you don"t have anything nice to say, don"t say anything at all", I swallowed my injured pride and kept going. In time, I might have realized the wisdom of my superior's decisions, and said a silent thanks to God. In any case, next time you are swayed by an image, particularly a representation of something attractive, as opposed to the actual thing itself, ask yourself one simple question. Is this true? Odds are, it is not.