Data journalism, as a new form of journalism, is gaining ever greater ground on newsrooms. Datasets and visualization applications have contributed to its bloom. Data and visualizations have a dominant role in the journalistic article, and their introduction affects the narration of the story, too. It is, also, considered by news organizations as a tool which provides information to users in meaningful ways and a part of the transition from being rather passive news-and-information sites to more news-and-information platforms. On data journalism projects we take into account the existence or absence of interactivity related to the visualizations and the amount of text that is included. The taxonomies of data journalism that have been proposed, take into account various parameters such as the content of the articles, the type of visualizations and the type of interactivity. In our research, we focus on a previously proposed taxonomy and we investigate how it applies to some of the most popular news organizations that develop online, data-driven, interactive news articles. By collecting data journalism articles, we try to check how visualization affects the story and the type of journalism interactivity. In particular, we examine the websites of The Guardian and The New York Times, which have been characterized as the "elite" of newsrooms, the British Broadcasting Corporation (BBC) and the Cable News Network (CNN), two of the most popular networks, and the Associated Press (AP) and Reuters, two of the largest news agencies.