“I have one of those jobs where my parents don’t fully understand what I do,” said Simon Rogers, the MRS Conference lunchtime keynote and editor of The Guardian’s Datablog.

Allow us to explain: Rogers is what you might call a data journalist. He creates stories out of data: whether that means writing about data or producing data visualisations.

The Datablog has become the biggest blog on The Guardian’s website, he said – which is odd, perhaps, given the weird relationship that exists between journalism and data. Playing around with numbers has “always been seen as a slightly odd, wonkish thing to do,” explained Rogers.

However data journalism has exploded along with the volume of data now being created, and with the number of tools available to help non-specialists do quite sophisticated things with that data.

Still, “a lot of the stuff we deal with is quite small data,” said Rogers. The Wikileaks logs might have ran to some 400,000 items – but that’s not really big data. “It seems enormous,” Rogers said. “You can’t possibly read through it all. But there are people out there who deal with millions of rows of data.”

But whether 400m or 400,000, Rogers said it’s always best to start by trying to make the data you’re working with smaller – to try to reveal the story within.

In picking what story to tell, however, Rogers said you can’t guard against bias. “I don’t believe in an objective truth,” he said. Being a news editor, he said, you get a lot of information thrown at you on a daily basis. “So you have to filter stuff out based on what’s interesting to you.”