I’m sure we’ve all heard some variation of the phrase “this is what the data says” or “the data speaks for itself.” It’s a nifty phrase, in that it conveys the idea that what is being “said” by the data is so basic and self-evident that it is uncontroversial, and that anyone that disagrees with the data “says” is debating reality itself. In my opinion, though, it is an insidious way to shut down debate. Data really doesn’t “say” anything at all.
I suspect that when many people picture “data” in their head, they’re picturing a pie chart, or bar graph, or some other graphical representation of some set of data. That’s not correct, though. This is what data actually looks like:
Haha, sorry, I just couldn’t resist that joke. To be honest, that’s half the reason I’m writing this post in the first place. That Data can talk, though.
Okay, seriously, here’s what data looks like:
What does this data tell us? Nothing, on its own. (Literally nothing, by the way, it’s just some random numbers generated in Excel.) How do we make data useful to us, then? By analyzing it.
It may seem like semantics to some people, but to me, the meanings of the words we use, and the ideas that they convey, are quite important.
Data is a set of observations. It is raw material, unspoiled and without meaning. We must conduct an analysis of data to generate any meaning from it. We look for patterns, we organize it, and we present our conclusions to others to convey the meaning that we find in the data.
This analysis, though, is what can cause problems. Any analysis is only as good as the person performing it, since people are fallible and biased. When conducting an analysis of a set of data, there are many ways to make a mistake or introduce bias. I’m no expert on data analysis, and my goal here isn’t to talk about the ways mistakes can be made; this article talks about some errors the article’s author considers common, though.
So don’t let anyone fool you into thinking that any data that isn’t a character on a late 80’s/early 90’s Sci-Fi television show can speak for itself. Some analysis must be done to find meaning, and that analysis can be misleading. Learn the basics of critical thinking, evaluate any analysis presented to you, and decide for yourself if the data backs up the conclusion the analysis makes.