Lies, Dam Lies, and Statistics. We have heard it so many times – mainly from people trying to make a joke to cover their own numerical discomfort.
I will admit that statistics can be used to lie. Sometimes the numbers are just completely made up. Sometimes, the numbers are intentionally misrepresented. The statistics are used to support the lie and intimidate opponents. We must only watch the nightly news to see examples. These situations cannot be blamed on the statisticians because, frequently, statistics are not truly involved.
Just as often, the numbers are wrong. Poor collection, incomplete methods, or incorrect analysis produces statistics that cannot be trusted. Even for the experienced analyst, it can take some time to deconstruct the inaccuracies. As I have said before, there are no perfect studies. Public presenters usually do not take enough time and instead rely on the reputation of the source. This problem is caused by the statisticians.
The bigger problem is when people quote statistics like they quote song lyrics.
Don’t go out tonight.
There’s bound to be a fight.
There’s a bathroom on the right.
For those of you that recognize the original song (Bad Moon Rising), you know how silly those words sound. Someone may joyously voice the new words while the meaning is lost to a message that sounds foolish.
Imagine a speaker trying to make a point based on the mangled lyrics. (Hard enough to understand the original lyrics.) The point may seem temporarily correct to those that don’t know the correct words. An effective and confident speaker can intimidate many to remain silent.
Just as the inaccurate lyrics are influenced by the person quoting them. The inaccurate statistics are influenced by the person representing them. The biases inside the head of the observer actually change the meaning of the numbers. The strength of the numbers are only temporary and the suggestion are misguided.
I once watched such a passionate person start a presentation with the “lies…statistics” quote. He then continued to quote statistics. In this case, he discussed a study that ranked the percent of children in poverty. He understood the original study to indicate that children in the United States were in worse conditions than in Bangladesh. The presenter jumped from the concept that there are more children in poverty in the United States than in Bangladesh to mean that children are generally better off in the impoverished country. He failed to understand that the percent in poverty was relative to the country’s income. The study’s ranking was affected by a greater income variance in the USA combined with a reduced median income in Bangladesh.
What sounded like an amazing point was quickly refutable. Children in poverty is a tragedy, but the problem is exasperated by inaccurate information. Eventually, people would believe that either US poverty statistics are inflated or that we no longer need to help the people of Bangladesh because they are doing better than America.