Making Sense of the Numbers: When to Embrace Relativity, and When to Ignore It

→
Making Sense of the Numbers: When to Embrace Relativity, and When to Ignore It

You pull up a chair to your desk, and open up the brand new data file you’ve just downloaded from SurveyMonkey. Ah. Nothing like the smell of new data! Anyway, you open up the file, and there you see……all kinds of numbers – 56%, 40, 98%, 479. What on earth are you supposed to do with them? When are you supposed to use the raw numbers (e.g. 40 and 479)? When are you supposed to use the percentages (e.g. 56% and 98%)? Are they even measuring the same thing? As your panic level rises the thrill of new data subsides and instead you just feel overwhelmed. Well, good thing you’re reading this blog post then!

The best and easiest way to start your data analysis is to look at the raw number of responses for any given question–they’re literally a count of all of the people who checked any given response. In research terminology these raw counts are known as “frequencies.” Examining your frequencies will always give you the most basic picture of your data. Okay, so let’s try an example. You look at your frequencies, and there are 987 people who think desk calendars should be outlawed. That seems like a lot. “Down with desk calendars!” you shout.