Brain Games: Accurate but Not Significant

Digital marketers are always changing, just like the industry itself. We test ideas, learn from those tests and, hopefully, do better next time. In the same fashion, we make decisions based off our own tests in the form of personal experiences, even though those experiences may actually misinform us.

Cognitive biases are human errors in how we think based on personal experiences and shortcuts we’ve developed. When it comes to data and analytics, cognitive biases can play a role in our interpretation of the data and how we tell the story. For example, the “Insensitivity to Sample Size Bias” states that humans don’t reliably dismiss data that comes from small samples. In other words, we’re not great at recognizing that smaller sample sizes don’t accurately reflect an average.

When we hear, "Four out of five dentists recommend Toothpaste X," we assume Toothpaste X is the way to go. In reality, this recommendation doesn’t carry much weight because the sample size isn’t statistically significant, even though the data is accurate. Digital marketers may know this, but as consumers, we quickly forget.

It’s important to recognize cognitive biases so that we continue to meaningfully interpret data and create a positive experience for customers, always keeping in mind that human nature plays a part in our digital decision-making.