Hey what’s up Thinkers! Kathy Gibbens here…
And, as I usually do, let’s start off by reviewing a fallacy we covered earlier this season: DoubleSpeak. Want to hit pause real quick and see if you can remember what DoubleSpeak is? Ok, so DoubleSpeak happens when someone purposely uses vague & ambiguous language to confuse or distract you from the real meaning of what they’re saying.
If someone is using DoubleSpeak with you and it’s sounding confusing, the question to ask them is this: “Can you explain that to me in plain English as if I were a five-year-old?”
If you want to review or hear more about this fallacy, go back & check out Episode 147.
Guys, if you’re needing a fresh go at math in your homeschool, check out CTC Math! Here's what Amber, a homeschooling mom, says about this online program: 'It's a rare find! My six kids, ages 6 to 15, find it easy and engaging. As a mom, I love seeing them grow in their math journey.' Intrigued? Check out ctcmath.com to learn about a program that families love.
Alright, today we are wrapping up our miniseries on Statistical Fallacies. I really hope this miniseries has been helpful to you! The reality is that it’s easy to be fooled & manipulated by numbers and data, especially when someone knows how to present them in a way that’s believable and compelling. But it can lead us to make false conclusions and believe wrong things, so we need to have the wisdom to recognize when someone may not be using data correctly, either on purpose or by accident. I started this little miniseries talking about the book titled How to Lie With Statistics, and I’ll end this series the same way. This short little book is a lot of fun to read and will definitely teach you how to analyze and recognize statistical fallacies when you see them! I think it should be part of every high school curriculum, it’s such useful information!
So let's dive into part 9, and talk about Data Dredging. Data Dredging happens when someone looks at a set of data and seeks out and tries to create more connections or correlations than are actually there. It’s also sometimes called Data Fishing or Data Snooping.
One example I found when I was researching this fallacy is a graph that I’ll send out today via email to those of you who are on my email list. And yes, that’s a shameless plug for getting on my email list! Anyways, some guy used Data Dredging to create a graph that shows the correlation between the number of letters in Scripps National Spelling Bee's winning word and the number of people in the United States killed by venomous spiders. Just by looking at this graph, it would appear that these two things are related to each other! Imagine the headline I could create from this graph: “Data shows winning spelling bee words accurately predict snake bites”! Of course, it would be ridiculous, but I could say “But the data proves it!” even though there’s not really a correlation.
The problem behind Data Dredging is that in our society, we’ve been so trained to believe that if someone has a statistic for something, then it must be true. And while, yes, statistics can and do tell us information, we have to know that they can also be manipulated in a way that tells us information that just flat isn’t true or that a correlation exists where it doesn’t actually exist.
A true and somewhat famous example of this statistical fallacy happened when a science journalist named John Bohannan created an intentionally false and misleading study in 2014 about how eating chocolate bars everyday can help you lose weight. It was called "Chocolate with high Cocoa content as a weight-loss accelerator". In this fake study, he intentionally used bad statistics and dredged the data so that it “showed” an end result that was totally false. He created this little experiment on purpose to show how the media would pick up & run totally meaningless studies, this one specifically around health & weight loss. He even created a fake institute called “The Institute of Diet & Health” and wrote a fake press release showing tantalizing pictures of women eating chocolate to go along with it. And it worked. The study was run not only in the US media, but also in Germany, India and Australia. See how that works?
So the question to ask yourself when you see correlations that seem to be backed by data is this: “Is it really true that those things are related in that way?” *repeat*
Remember: When you learn HOW to think, you will no longer fall prey to those who are trying to tell you what THEY want you to think and it all starts with asking one simple question: “Is that really true?”
