• Start from a headline and finding data to support or refute it. It’s a lot easier than digging around in a dataset trying to find the story.
• Select a topic you’re passionate about, and go from there. That way when your data doesn’t work out you have secondary ideas or data sources you can pull from.
• Don’t start from a dataset - if it starts going poorly, you’ll probably fail to acknowledge it and drown in misery as a result. Not all data is your friend!
• Failure is an option! Sometimes things just don’t work out, and as long as you write it all up and learn something you’re doing great.