Falsified Data

Originally Posted May 4, 2017

---

Our world is full of false data. One of my first posts, Muslims to Halal, contains an example of this. “Fake News” has appeared everywhere, blooming after Donald Trump’s claimed anything he didn’t like the sound of as “Fake News” and irrelevant (with no evidence to contradict it). Even the controversial vaccines causing autisms debate, that we now know was based off a completely fabricated study, still manages to appear in arguments against vaccines anywhere you look. This is so obviously false, yet people still believe it. How can we possibly tell the false data from the real stuff with so much bombarding us but so little of the truth? Everywhere you look, something can be questioned, challenged or disregarded.

The idea for this blog post started when one of my lecturers mentioned something of interest. He pointed out the efficiency of the carbon cycle in agriculture. I’ve always heard that cows produce the most methane (carbon = bad) of pretty much anything, ever. This knowledge is used by numerous people, companies and governing bodies for multiple reasons, one major one being the pollution of other carbon producing processes. Why blame coal for carbon emissions when there’s a huge paddock of cows producing the same stuff?

What the lecturer suggested, though, was the missing information. Cows fart out all that carbon, but they’re standing in a paddock of plants that use carbon like we use oxygen. The oxygen produced by the cows is absorbed by the plants, which then grow and are eaten by the cows to be farted out all over again. I believe this is what he meant by the efficiency of the agricultural carbon cycle. In a coal plant, surrounded by concrete and parking spaces and factory buildings, plants are far off in the distance and not present to absorb the dangerous fumes that are then added to our atmosphere to contribute to global warming (if you believe in global warming, that is).

Yes, cow farts may contribute hugely to the earth’s output of carbon emissions, but this carbon is quickly recycled and unlike other potential emission sites may not have such a horrible effect on global warming.

In another class I learnt a few interesting things about falsified data in pharmaceuticals. This was during a talk by Ben Goldacre in “What Doctors Don’t Know About the Drugs They Prescribe”, though I can’t remember what facts came from him and which came from our general discussion afterwards (sorry!). One point made was how pharmaceutical companies can rig data using placebos. Without placebos, most experiments are pointless. If you don’t know what a placebo is, most common example is the sugar pill. Giving one group pills with the drug being tested, and another group pills with simple sugar and no drugs (but the group still believe they are receiving the drugs), allows you to compare the people with the drug to the people without, and eliminates belief (when people get better simply through believing the pills they take are working). In this effect, the sugar pills are the placebo.

In drug trials, sometimes companies can claim their drug works. But testing a drug against a placebo that is simply sugar pills or water (so, nothing), is pointless when there’s already a drug in effect that also ‘works’. What companies should do is test the new drug against the original drug to see if it is better than the current drug of choice. Companies can claim their drug works by avoiding this through choice of placebos, and still claim their drug is tested to work in general (and therefore still gain a profit even if their drug isn’t as good as the first was anyway).

There was also the case of withholding data. The example Goldacre used was flipping a coin. If you flipped a coin 100 times, you’d expect (about) 50 heads and 50 tails. Now imagine each of these was a different experiment. You got 50 positive studies for your product, and 50 negative studies for your product. If you withhold 50% of your studies, you can technically say your product was 100% positive (of the studies you released). The same as saying your coin had two heads and no tails, because it always landed on heads, so how could it have tails if it never landed on tails?

The withholding of studies in such a way is apparently quite common in the industry, and a huge problem when trying to determine the truth. Even the doctors that want to research the absolute best options for the patient may find nothing wrong with a product. Yet there could be many studies that showed the product as risky, but only the best studies showing the product in a positive light were published and made available for the doctor to read.

There is so many ways data could be falsified that it seems impossible to tell fact from fiction. Key points I came across in both these classes was a) methods b) data and c) intentions of those producing the studies. No matter what, there always seems to be bias, I think total unbiased studies will never truly exist, but aiming for the least bias is always a good start. Methods, often unseen by the public, hold key points that determine how relevant the data is. For example, side effects of drugs were studied to produce the drugs with the least side effects, BUT these were only ever tested on males (until recently I believe). As a result, women often suffer unpredicted side effects of many common drugs that can claim the side effects due to something else due to the amount of studies showing side effects aren’t caused.

Another example used that I can’t quite recall followed these basic concepts: A study was performed with a series of questions. One of the questions asked, “Do you associate yourself with Jihadist principles”. A scary number said yes. Another question further down asked “What do you believe Jihadist means”, and a certain number associated jihadists with the terrifying war many of us see it as, but a few (some of which were those that said “Yes” to being Jihadist) defined it as more peaceful inner striving to best believe in their religion. So even though statistics say a large percentage of jihadists live in America, it failed to explain that those jihadists may not be terrorists but just devoted to their religion in a peaceful manner. The scary statistic would be those that are jihadists AND associate jihadists with war and death.

I truly do not know how to take all this data and tell wrong and right. I don’t think it will ever be that simple, with a million shades of grey versus black and white. There is always some truth in everything, and some things that others would take as incorrect. It’s an impossible struggle, but something we must strive for in order to trust and understand the abundance of knowledge around us.

Share your tactics for trusting information, or deeming it as “Fake News”. Share your stories of fake news, and how data was misinterpreted below in the comments section, or if you don’t wish to comment feel free to contact me.