Here’s another snippet from my data journalism book chapter (now published). As part of my research David McCandless, author of the very lovely book and website Information is Beautiful gave  his 5 tips for visualising data:

1. Double source data wherever possible – even the UN and WorldBank can make mistakes
2. Take information out – there’s a long tradition among statistical journalists of showing everything. All data points. The whole range. Every column and row. But stories are about clear threads with extraneous information fuzzed out. And journalism is about telling stories. You can only truly do that when you mask out the irrelevant or the minor data. The same applies to design which is about reducing something to its functional essence.
3. Avoid standard abstract units – tons of carbon, billions of dollars – these kinds of units are over-used and impossible to imagine or relate to. Try to rework or process units down to ‘everyday’ measures. Try to give meaningful context for huge figures whenever possible.
4. Self-sufficency – all graphs, charts and infographics should be self-sufficient. That is, you shouldn’t require any other information to understand them. They’re like interfaces. So each should have a clear title, legend, source, labels etc. And credit yourself. I’ve seen too many great visuals with no credit or name at the bottom.
5. Show your workings – transparency seems like a new front for journalists. Google Docs makes it incredibly easy to share your data and thought processes with readers. Who can then participate.