Journalist Paul Bradshaw says data journalism is “incomprehensibly enormous,” in part because it represents the convergence of several fields — programming, design, statistics and investigative research, to name a few.
Bradshaw wrote about this in a Guardian story that’s worth revisiting. He simplified data journalism by separating it into four parts: finding data, interrogating data, visualizing data and mashing data. Bradshaw recommended that journalists start off with a question and then find the necessary data to answer the question and ultimately develop a story.
Often, Bradshaw said, journalists need to contextualize the initial data they find with further data, and they need to clean up data by fixing spelling errors and poorly formatted fields. Tools such as Freebase Gridworks can help with this.