Understanding Bias In Clinical Research

Why reading research can be much more complicated than you would think

An AI representation of ‘bias’. Deeply weird, but also not that terrible. Photo by Google DeepMind on Unsplash

Reading clinical research is quite difficult. There are a lot of different ways that you can be mislead when you come across a study, because the things that make a piece of research reliable or not are finicky and diverse.

But there’s an overriding idea about most of the stuff that makes research unreliable. In the epidemiological world, we call this bias. There are many different ways to define bias - it’s a complex concept - but the one I like the most is any issue in design or implementation that reduces our confidence that the question asked by the study has been answered correctly.

That may sound nebulous to you, but don’t worry. The entire point of this piece is to help you understand why bias is important, and why it’s important to at least know that you can be misled by bias in trials even if you aren’t always aware that it’s there.

Bias is any issue in design or implementation that reduces our confidence that the question asked by the study has been answered correctly.