Excerpt for Truth, Lies & Statistics by , available in its entirety at Smashwords

Truth, Lies & Statistics





or



How to Lie with Statistics















Copyright



Truth, Lies and Statistics

By Lee Baker

Copyright 2017 Lee Baker

Smashwords Edition



Thank you for downloading this ebook. You are welcome to share it with your friends.

This book may be reproduced, copied and distributed for non-commercial purposes, provided the book remains in its complete original form.

If you enjoyed this book, please return to your favourite ebook retailer to discover other works by this author.

Thank you for your support.











Contents



The Uncomfortable Truth About The Truth

Are You Biased?

The Average Human Being Has Only One Testicle

Give Ps A Chance

The Confidence Trickster

Pay No Attention To That Man Behind The Curtain…

Pirates Caused Global Warming

The End (Nearly)



About the Author

Claim Your FREE eBook Now!

Leave a Review











Testimonials

As this book is hot off the press, it hasn’t had time to gather any reviews or testimonials yet. Not to worry – when it’s got some lovely reviews I’ll add them here.

So chop chop – if you get a move on and leave a juicy review you might end up getting a mention right here!

I’ll remind you at the end of the book to leave a review or testimonial.



For now, enjoy!











Claim Your FREE eBook Now!





This is the sister book to Truth, Lies and Statistics, and shows you how to lie with graphs

(if you’re unscrupulous enough…)



Download your FREE copy right here:

https://chi2innovations.lpages.co/book-graphs-don-t-lie/











The Uncomfortable Truth About The Truth

In the immediate aftermath of Donald Trump’s inauguration as President of the United States, it was claimed angrily and repeatedly that the size of his inauguration crowd was the “largest audience to ever witness an inauguration – period…”. White House staff backed this up by quoting the numbers of people that rode on the Washington DC Metro on that day and compared it to the figures for Barack Obama’s inauguration.

What more proof do you need?

Well, for starters, we need these numbers to be quoted correctly (they weren’t). The official DC Metro ridership figures show that Trump’s inauguration crowd was actually the smallest of the last four inaugurations.

So, was it a lie? Well, it certainly wasn’t the truth, the whole truth and nothing but the truth.

You see, it is easy to fall into the trap of thinking that truth and lies are binary. If it’s not the truth, it must be a lie. If it’s not a lie, it must be the truth.

Not so. There is a grey area in-between that we all fall into from time to time, whether by accident or otherwise.

I’m sure you’ve heard of such terms as half-truths, partial truths, preferred truths, uncomfortable truths and alternative facts. Similarly white lies, fabrication, exaggeration, bias and deception. These all fall into that grey area.

Statistics also falls into that grey area.

Before I get an avalanche of emails decrying me as the spawn of the devil, let me explain.

Data doesn’t lie. People do. If your data is biased, it is because it has been sampled incorrectly or you asked the wrong question (whether deliberately or otherwise).

Statistics, on the other hand, does lie. When analysing data it is rarely the case that there is one correct approach and all other ways are wrong. There are often many ways of analysing the data, some of which are more appropriate than others, and these different approaches usually give different answers. Take as an example ‘the average’ – the central point of your data. There are over a dozen different ways of calculating it. Take a sample of data and work out all the different averages. Which one of them is correct? Most likely none of them. Some of them will be closer to the true central point of the data than others, but it is extremely rare that statistics gives you the truth and nothing but the truth.

By careful selection of a particular statistical method you can get a result that is close to the truth or very far from it, as is your wont. And this is what I mean by saying that statistics lies. It rarely tells the entire truth and you can make it as close to, or as far from, the truth as you like. If you’ve spent enough time around statistics, you’ll know that you can make your data say pretty much whatever you want it to, if you’re so inclined.

Of course, good scientists, researchers and statisticians wouldn’t dream of doing such a thing, would they? Or would they…

Have you ever heard the phrase ‘9 out of 10 cats prefer…’? Of course you have. After all, they are a multi-billion dollar company who has commissioned a hugely expensive advertising campaign to persuade us to buy their cat food rather than that of their rivals. And yet seemingly they can’t afford to have more than ten cats in their trial. There’s definitely something fishy here, and it’s not just the flavours of their cat food.

If the researcher works for a commercial organisation, the results of their analyses will always be subject to question because – even if their genuine, truthful and unbiased results are strong – artificially inflating the results can lead to a greater volume of sales. The company has something to gain.

Similarly if you work for a contract research outfit that is being paid by a company to do some research and analysis. On one hand if you produce a biased study you will leave the integrity of your company open to debate, but on the other, a future contract with the paying client might be at stake if the results don’t sparkle quite as brightly as they might. “More than 80% of dentists recommend our toothpaste”, claimed one advertising campaign. Does that mean that fewer than 20% of dentists recommend all the other toothpastes combined? Not necessarily, but that’s the clear implication. The data collected may have been correct, but was the survey designed to mislead? Almost certainly!

You don’t need to be creative with your choice of statistical analysis to deceive, though. Oh, no, we humans are hard-wired to look for patterns. Just show us a pretty little pattern in the data and we’ll make up our own story to explain it. Like how global warming has increased as the number of pirates has decreased. The obvious inference is that the lack of pirates caused the planet to warm. It’s true, honest, and there’s a new religion that’s sprung up around it. More on that story later…

Better still, we might even be able to use statistics to ‘prove’ that the pattern is a real one, like the correlation between autism and organic food sales. Does organic food cause autism, or is it the other way round? Similarly with the correlation between imported Mexican lemons and deaths on US highways, implying that Mexican lemons are killing Americans. There surely must be something in these because the correlations are highly significant.



I hope you can see that it is not possible to tell the whole, unvarnished truth all the time. You can’t even tell it most of the time. In fact, you might not be able to tell it at any time, but if you are honest about all sources of potential bias and weaknesses in your study methods, you will come out of it with more credit. And that’s never a bad thing.

On the other hand, if you’re comfortable with half truths and alternative facts – and even prefer them as a vehicle for profit or career enhancement – then this book will help you find lots of ways to tell untruths with your data.

A word of caution though: this book is not written for you. It’s written to help good, conscientious researchers spot when you’re trying to pull a fast one. For every conman there has to be a good, honest man ready to point an accusing finger.

After all, as the saying goes – evil flourishes when good men do nothing…





Back to Contents











Are You Biased?

The short answer to this question, as we have already established, is Yes.

The long answer is still Yes, but with the caveat that biases may be understood, controlled for and minimised – if indeed that is your intention – but not eliminated entirely.



Take the example of the airline industry. The United States airline company Boeing reported over 30,000 passenger fatalities in their 2016 annual Statistical Summary of Commercial Jet Airplane Accidents report. Wow – that is a huge number of deaths of American citizens in a single year.

Clearly it’s not safe to fly.

Except that it’s not true. What is true is that Boeing reported over 30,000 deaths in their 2016 annual report. By carefully choosing my words, I accidentally on purpose made it look like all these deaths occurred in 2016. They didn’t. In fact Boeing reported that these were the cumulative deaths from 1959 to 2016, and not just in the US, but worldwide. The average figure (for what it’s worth – it’s ridiculously easy to lie with averages, as we shall see later) is just 526 worldwide airline passenger deaths per year, and the annual accident rate per million departures has declined annually from 50 in 1959 to around 1 in 2016.


Download this book for your ebook reader.
(Pages 1-7 show above.)