The Data That Divides Us: Before Posting That Link, What Does It Really Say and How Is It Biased?

It is near impossible to transverse the digital world without experiencing the divide and discourse in our society. It seems almost every substantive article, social media post, tweet, or even comment on a site that deals with social interactions quickly devolve into political mudslinging. People will share the viewpoint of their side with “claims” that are often just made up or at best cherry-picked statistics to support their viewpoint (Confirmation Bias), often not looking beyond the title of the reference research or the abstract. Even when the opposing side has something that we could learn from we are so entrenched that we can’t or don’t want to hear it (Ostrich Effect) and instead resort to name-calling using adjectives like “Radical” “Extreme” or “Far”. For instance, we have a large percentage of our population (35% based on a recent NPR-Ipsos Poll) who believe that Trump won the 2020 election, but was denied the win due to massive fraud, which has been proven false and failed in over 60 court cases.

So, what should we do about this? Maybe turn to a book….

I recently finished reading “The Data Detective” by Tim Hartford (Titled “How to make the world add up” outside the US). If you’re not familiar with Tim Hartford, he is a UK-based economic journalist, author, podcaster, and most excellent storyteller.

The book layouts 10 “commandments” to help us understand the data, charts, and statistics that bombard us every day. Hartford weaves stories and examples into each commandment to explain and contrast the lessons.

The 1st lesson is the most important: Check your emotions; How does the news article, or chart makes you feel, how can you stop to think, and consider if you are missing key data because of your emotions and bias’. Now if you read this and say, “I’m not biased – only the other side is”. Well, I’ve got news for you, we’re all biased. If you’ve ever applied a heuristic (a mental shortcut) you’ve applied bias. If you’ve ever judged a group of “others” (A different religion, different nationality, political group, skin color, etc.) or assumed a generalization on that group YOU ARE BIASED. If you can’t see your bias, you’re not looking!

The other 9 rules cover, things like, (2) exploring how your personal experience may be different from the statistics. (3) Understanding what’s been measured and seeing if they are compatible. (4) Putting the statistic into content– does it make sense based on other knowns – can it be verified. (5) Where does the data come from and (6) what data may have been excluded, either unintentionally or maliciously. (7) How big data regression can lead you astray – especially if they aren’t open to interrogation. (8) Pay more attention to the bedrock of official statistics. (9) Pretty graphics can hold good and bad data, Check that you understand the basics of the chart. What does the axis mean? Does the title match what the data says? (10) Keep an open mind, asking how we might be mistaken and whether the facts have changed.

When you next see a social media post that supports, your view before retweet-post-sharing, check the source of the data for fairness and accuracy. Equally, if you see a post to something you don’t agree with start by asking why you have that emotional response, and how could your viewpoint be wrong? And please please please, if you state a claim in a post, comment, or blog, link to a reliable robust, and least biased source, and check that your viewpoint is supported by the data, not the other way around.

The Data detective can be found in bookstores or online. Alternatively, a summary of the book is available on Tim Hartford’s podcast “Cautionary Tales” on both Apple Podcast and Google Podcast.