In an article on The Next Web, developer and writer Mina Demian compares data journalism to basic science, because in both cases you want to use numbers and stats to prove or refute a statement with scientific evidence.
Demian outlines the process for data journalism beginners, from how to structure your work to the step-by-step process, providing a real-world case study about unemployment in South Africa in 2015 to explain it better.
The original data contained more than 20 spreadsheets but he explains how to summarise only the information actually needed for your story in just one spreadsheet.
The next step is to visualise all the data that has been collected. When creating a chart, make sure you explain what the numbers mean and “always aim to show the values on the chart, where appropriate” as it makes it easier to understand the graphics right away.
However, you're not done yet – find experts or analysts to interpret and comment on your data, and compare it to other datasets to add context and test your findings before writing and publishing your article.