Methodology: multilevel data collection
The research tools used in this study were collaboratively developed in order to guarantee a maximum degree of intercultural validity. Altogether 21 countries participated in the pilot study: Australia, Austria, Brazil, Bulgaria, Chile, China, Egypt, Germany, Greece, Indonesia, Israel, Mexico, Pakistan, Portugal, Romania, Russia, Spain, Switzerland, Turkey, Uganda and the United States. In each of these countries we conducted interviews with a quota sample of 100 working journalists drawn from 20 news organizations. Interviews were mostly completed between October 2007 and January 2010 by telephone or face-to-face. In addition to the interviews we also collected data on their news organizations and on relevant properties of their media systems.
Sampling was carried out in two steps. We first selected 20 news organizations in every country according to a common quota scheme (see Table). Wherever possible we selected five journalists in each newsroom. Journalists were defined as those who had at least some “editorial responsibility” for the content they produce. From all 413 newsrooms that were chosen in the first place, 28 refused to cooperate and were subsequently replaced. On the level of the journalists, we had to substitute 266 interviewees from the altogether 2100 journalists due to refusal.