One of my friends often says, “About 72% of statistics are made up on the spot.”
There’s no science to back up his claim, and he doesn’t always use 72 as his number. Still, I can’t help but think there’s a lot of truth is his statement.
For the last few weeks, I’ve been reading a lot about new year’s resolutions. I’ve seen stats saying 50–90% fail in the first week, month, or quarter of the year.
Who’s making up that number?
Even if the stats are poll-related, how can they be accurate? How many people are honestly admitting they had resolutions or how long they kept them?
Doctor House’s voice rings in my ears, “Everybody lies.”
Am I going to tell you how long I really stuck to my diet resolution? I might push the poll button saying I made it three weeks, because that’s how long I’ve read other people lasted.
In reality, I probably lasted 12 hours (curse you, Krispy Kreme!), but I’m too embarrassed to enter that in a poll.
Make Your Own
If statistics are unreliable, which, let’s face it, in today’s news and social media posts, they often are — not that we’re pointing any fingers.
You know who you are.
Anyway, if statistics are dubious, then let’s all make up our own.
Right now, 29% of you disagree with me. Another 32% think I’m right but would never admit it in public.
There’s 48% of you that haven’t even read this far but will still clap for this story.
Thank you, I guess.
Another 56% had their eyes glaze over as soon as they started seeing numbers and the percent sign.
By now, 81% are thinking about something else. A small 3% are wondering if any of these stats are accurate, while 66% are thinking about what to eat next. More than 10% of you just farted, and another 23% are picking your nose.
Stats are funny things. Unless we base them on clear scientific empirical data, most statistics have no value.
For example, imagine you want to publish statistics about people who like chocolate. You know five people that hate chocolate.
I know, it sounds insane, but there really are such strange creatures in our world.
If you only interview those five people, your data will show that 100% of people don’t like chocolate. If you throw in five people who like chocolate, your stats go to 50%.
I’m not saying that’s what happens in every poll, but come on, how many polls do you imagine and entirely unbiased? Should we guess 18%?
We all know where to find the people that think (or don’t think) like us. Pollsters are no different.
In the tax world, we have to deal with a lot of ceiling numbers. Are you wondering what ‘ceiling numbers’ might be? Here’s a scenario.
The tax preparer starts, “Mr. McGillicuddy, I see you drive your vehicle for business purposes. How many miles did you drive last year?”
Mr. McGillicuddy looks at the ceiling. “Ah, about 7200.”
Our tax preparer continues, “Mrs. McGillicuddy, I see that you sell cosmetics. Your sales came in at $5000. How much did you spend on supplies?”
Mrs. McGillicuddy finds the same spot on the ceiling. “I spent $3200,” she says.
I suspect that most statistics are in reality ceiling numbers.
There are 37% of you mad at me right now. You live and die by polling data and refuse to believe there is any margin for error.
On the other side, 84% of you think I’m right. You’re nodding your heads, and 3% of you are even giving me a standing ovation.
Yes, those numbers add up to more than 100, but does it matter if you’re making them up?
Right now, 32% of you are smiling as you recognize how senseless statistics can be. About 9% stopped reading.
That’s 78% okay.
In closing, I want you to consider the next set of statistics you see. Are they accurate or part of the 72% pulled from thin air? Do they reveal true findings or are they 87% biased? Are they ceiling numbers?
To be honest, I 99% don’t care.
Scott Ninneman is a bookkeeper by day and a writer by night. He’s the voice of the blog, Speaking Bipolar, and writes about living with bipolar disorder and chronic illness. He also enjoys writing short stories, poetry, and inspiration for personal development. His interests include reading, cooking, and entirely too much TV.