This paper is the essence of the first training School of Data Journalism which was held European Journalism Center, Open Knowledge Foundation, and International Journalism Festival. This session was hosted by Steve Doig, Knight Chair in Journalism, expert in the field of Computer-asisted reporting
Why do data-driven coverage
Doig believes that data-driven journalism enables journalists to go beyond anecdotes and rely on facts and evidence. Using the data to do coverage, one can see the best point - the most illustrative thing about a particular story. So how do you come up with ideas for data-driven reporting? First, take a look at the various coverage you've done such as sports, elections, disasters, criminal investigations, cash flow, and so on. Almost all subjects covered by journalists have data that can be analyzed in more depth. You, too? can find ideas from various sources such as:
- See what other journalists have done before? If something is happening in a city, it is possible that it is also happening in your city.
- See the various coverage that has been done at DataDrivenJournalism.net
- Check IRE's Extra feed
- Check out The Guardian Datablog
- Reading documents containing various data such as those created by State institutions and academics. Paying attention to footnotes and bibliographies can lead to interesting data sources.
Where do you get story ideas?
Re-examine your ideas by:
- The statement you want to make. Start with a hypothesis like? Crime is getting higher in my area.? From this hypothesis you will probably make various follow-up statements such as crime increased by a percentage, the crime rate per 1000 population in this area is the highest compared to other areas.
- Think about the variables needed to strengthen the statement. Start thinking like a table of information (columns are variables and rows are individual data). There are two forms of variables, namely:
- Categorical: gender, type of crime, zip code or any variable with a label.
- Numerical in character: numbers, number of crimes, number of accidents, number of detentions.
- Find out who has the data you need. After creating the desired variables, find out who owns the data. Institutions and organizations such as governments, corporations, and so on, collect a lot of information so you don't have to collect it yourself.
- Get the data. Don't be intimidated by the format of the data you can get. If you already know what to do with the available data, the format of a data can be changed to another format as needed by asking for help from someone who is an expert in this matter. You can find people who can do it on various mailing lists such as:
- Clean up data. Often the data is messy. For example, a report on campaign funds carried out by volunteers with various typos in it. This must be fixed first. Some software that can help in this process include:
- After the data is cleaned, what should be done? Look for patterns! High, low, maximum, minimum, and so on. Think about the shape of the data, look for anything that seems out of the ordinary. Keep in mind that many stories are discovered by simple things like sorting data. You can use some simple tools on spreadsheets such as sort, filter, function, and pivot tables.
The thing to remember is that data-based journalism is better done as a team. There are many different roles to play such as reporter, editor, graphic designer, photographer, videographer, web designer, and so on.
translated and extracted from?From Idea to Story: Planning the Data Journalism Story?