How to Protect Yourself From Being Manipulated by Data

In ancient Egypt, it was widely believed that scarabs, amulets designed in the shape of a beetle, had supernatural properties and granted good fortune to the wearer. In medieval Europe, horseshoes were also believed to have special powers and people hung them above doorways to attract good luck and to repel evil spirits. I could go on with more examples, but I think you can already see the pattern: throughout history people have attributed magical powers to ordinary objects. When we talk about these beliefs now, people sometimes chuckle or poke fun at how gullible people living in ancient times were. The irony is that this tendency hasn’t really gone away even if it gets expressed in a different form. And there’s one specific form that I’ll be focusing on today: data.

In today’s world there’s a certain fetishization of data where it’s perceived as having an infallible, larger-than-life stature, much like the fetishized objects of yore (the original meaning of “fetish” was a “material object regarded with awe as having mysterious powers”). If you make a claim and you have some data to back it up, why, then it must be true. But how is this similar to an object having magical powers? Isn’t data scientific and objective? It certainly can be, but it all depends on how the data is selected, collected, presented, and interpreted, and it’s rare that all of those things are done with equal objectivity. Unfortunately, just as with the snake oil purveyors of the ancient world, there are those who misuse data to manipulate others. Fortunately, you can protect yourself in the same way, through critical thinking or rather a very specific form of critical thinking that we can call data literacy.

Data Can Be Objective but People Often Aren’t

Let’s be clear about something. As an academic, I value data greatly. This is in no way one of those “science-is-fake” type sentiments you sometimes see on social media. To be clear, it’s not that science is fake or completely subjective. It’s that data, which is a product of science, can be weaponized and misused to manipulate you to think or behave in a certain way.

In addition to being an academic, I’m also a former trial lawyer. In law we have a legal equivalent to data; it’s called “evidence.” But here’s the thing about evidence: although two opposing sides are operating under the same laws and procedures, how each side presents their selected evidence invariably leads to opposite conclusions. Data (or “evidence”) outside the courtroom can work the same way. It can be (mis)construed to support different conclusions based on how one selects and presents it. A common example of this is shrinking or blowing up a graph to make an increase or decrease look more significant than it actually is.

So, the point I’m making is that data is indeed very important, so important that you need data literacy to avoid being manipulated by it.

The Principles of Data Literacy

How, then, do you cultivate data literacy? As with most topics on this blog, there’s no easy universal formula. But here are some principles you may wish to adopt.

1. Develop a general foundation of critical thinking

Data literacy is a subset of critical thinking, and so the first step is to become a better critical thinker in general. (Fortunately, I have no less than three posts on this blog that cover various dimensions of critical thinking, including why it boosts your ability to influence, common pitfalls for critical thinkers, and using critical thinking to understand the narratives that drive people.)

2. Ask yourself who stands to benefit and how?

This one is fairly easy to do, with a common and relatable example being when someone is literally trying to sell you something and directs a mass of data your way to support their case. Clearly, they stand to benefit from your buying whatever they’re selling. Now, it doesn’t necessarily mean that they’re deliberately misrepresenting the data in an effort to manipulate you. But they do have a clear stake in your being persuaded so it pays (often literally) to be on guard.

3. If it’s specialized knowledge, who can you ask about it?

There are situations where someone could be presenting an argument with data that allegedly supports it, but you’re not an expert or highly informed on the topic so you’re not sure what to make of it (or at least that’s how you should feel if you’re not informed about something). You can always ask someone who is informed on the topic what they think, just don’t forget to maintain your critical thinking perspective. You are not obliged to reflexively accept the views of someone who may know more about a particular topic than you do. This is one reason, for example, why people often bring an experienced car buyer with them when they go to a car dealership.

4. Ask yourself what’s missing from the data?

Sometimes people present data or evidence in a way where they’re not overtly lying, but rather lying by omission. This happens in the selection process when they leave out one or more critical pieces of information that may make all the difference.

For example, let’s say a tire salesman is trying to sell you an extended warranty and gives you all these facts and figures about how much money you would save by buying one, along with dazzling you with how much other customers have saved with their extended warranties. On the surface it may sound like an obvious choice, but what data is being left out? It could be a lot of things. Maybe, for example, the percentage of their customers who have actually had to use their extended warranties is just 1 percent, in which case the chances of your ever actually needing that warranty are extremely low. So always ask yourself, what’s being left out?

Cognition Essential Reads

How to Shift Your State of Mind

The Psychology of Disliked Music

5. If something sounds too good to be true, there’s a good chance that it is

This time-honored truism also applies to data, and it often goes hand-in-hand with principle #3. Sometimes people will stack multiple pieces of data in an effort to build what looks like an irrefutable case for something. To again use our example of the salesman and the extended warranty, the more all the data makes something sound like the only reasonable choice, the more a bit of healthy skepticism is warranted. The company wouldn’t be selling extended warranties unless it ultimately worked out in their favor (principle #2). This is because the majority of people who purchase extended warranties end up not actually needing them.

6. Be wary of your own confirmation bias

It’s not just other people’s biases that you have to be careful about. You have to be careful about your own as well. We have a natural tendency to accept data uncritically, and a particular interpretation of that data, if it reinforces a position we already hold. This is called confirmation bias, and the best way to start protecting yourself against your own confirmation bias is to notice when it happens — and believe me, it happens a lot. When you notice yourself wanting something to be right, try to intentionally think of how it might be wrong. You might in fact be right, but this way you’d be reducing the chances of being in error.

By no means do these principles cover everything about data literacy, but they are a good start. And none of this is to say that people in these situations are always manipulating the data or deliberately trying to fool you. Even if they have something to gain from your being persuaded, maybe their argument really is sound. And even if they are misrepresenting the data, it doesn’t mean they’re necessarily doing it on purpose. Maybe they misinterpreted the data or fell for their own confirmation bias. So, the point isn’t to automatically reject people’s arguments in these situations, and, conversely, these principles don’t guarantee that you won’t get manipulated. The critical thinking associated with data literacy simply reduces the chances of your being led astray, and in today’s data and information-saturated world that’s a goal worth pursuing.