²Oguz B. Nayman, "Professional Orientations of Journalists: An Introduction to Communicator Analysis Studies," Gazette 14 (1973): 195-212.

³Sven Windahl and Karl Erik Rosengren, "Newsmen's Professionalization: Some Methodological Problems," Journalism Quarterly 55 (1978): 466-473.

⁴Gertrude Joch Robinson, "Fünfundzwanzig Jahre Gatekeper-Forschung: Eine kritische Rückschau und Bewertung," (trans.: "Twenty-five Year Gatekeeper Research: A Critical Review and Assessment") in Gesellschaftliche Kommunikation und Information, eds. Jörg Aufermann et al. (Frankfurt: Athenäum, 1973), 345-355.

⁵George A. Bailey and Lawrence Lichty, "Rough Justice on a Saigon Street: A Gatekeeper Study of NBC's Tet Execution Film," Journalism Quarterly 49 (1972): 221-229.

⁶David M. White, "The Gatekeeper: A Case Study in the Selection of News," Journalism Quarterly 27 (1950):383-390.

⁷Ulrich Hienzsch, Journalismus als Restgröße. Redaktionelle Rationalisierung und publizistischer Leistungsverlust (trans.: Journalism as a Remainder: Editorial Rationalization in the Newsroom and the Waste of Journalistic Potential) (Wiesbaden: Deutscher Universitäts-Verlag, 1990).

⁸David H. Weaver, ed., The Global Journalist: News People Around the World (Creskill, NJ: Hampton Press, 1998).

⁹Wolfgang Donsbach (1981): "Journalisten zwischen Publikum und Kollegen. Forschungsergebnisse zum Publikumsbild und zum in-group-Verhalten," (trans.: "Journalists Between Audience and Colleagues. Empirical Results on Image of Audience and on In-group Behavior of Journalists") Rundfunk und Fernsehen 29:2-3 (1981):168-184.

¹⁰S. Robert Lichter, Stanley Rothman and Linda S. Lichter, The Media Elite: Americas New Powerbrokers (Bethesda, ML: Adler & Adler, 1986).

¹¹Herbert J. Gans, Deciding What's News: A Study of CBS Evening News, NBC Nightly News, Newsweek, and Time (New York: Vintage Books, 1980).

¹²Siegfried Weischenberg, "Die Unberechenbarkeit des Gatekeepers. Zur Zukunft professioneller Informationsvermittlung im Prozess technisch-ökonomischen Wandels," (trans.: "The Uncertainty of the Gatekeeper: On the Future of Professional Information Dissemination in the Process of Technological and Economical Change") Rundfunk und Fernsehen 33 (2, 1985): 187-201.

¹³Niklas Luhmann, Soziale Systeme. Grundriß einer allgemeinen Theorie (trans.: Social systems: Outline of a General Theory) (Frankfurt: Suhrkamp 1984), 250. Niklas Luhmann, Social systems (Stanford, CA: Stanford University Press 1995).

¹⁴Michael Fleischer, "Concept of the 'Second Reality' from the Perspective of an Empirical Systems Theory on the Basis of Radical Constructivism," in Systems: New Paradigms for the Human Sciences, eds. Gabriel Altmann and Walter A. Koch, (Berlin, New York: Walter de Gruyter, 1998), 423-460. Jouko Seppänen, "Systems Ideology in Human and Social Sciences: History and Philosophy of Systems and Model Thinking, Information Theory and Cybernetics," in Systems: New paradigms for the human sciences, 180-302.

¹⁵Siegfried Weischenberg, Martin Löffelholz and Armin Scholl, "Journalism in Germany," in The Global Journalist: News People Around the World, ed. David H. Weaver (Cresskill, NJ: Hampton Press 1998), 229-256.

¹⁶John W. C. Johnstone, Edward J. Slawski and William W. Bowman, The News People: A Sociological Portrait of American Journalists and Their Work (Urbana, Il: University of Illinois Press, 1976).

¹⁷David H. Weaver and G. Cleveland Wilhoit, The American journalist: A Portrait of U.S. News People and Their Work (Bloomington, IN: Indiana University Press, 1986). David H. Weaver and G. Cleveland Wilhoit, The American Journalist in the 1990s: U.S. News People at the End of an Era (Mahwah, NJ: Lawrence Erlbaum Associates, 1996).

¹⁸Armin Scholl, "Sampling journalists," Communications 21:3 (1996): 331-343. Siegfried Weischenberg, Martin Löffelholz and Armin Scholl, "Journalism in Germany," 232 ff.

¹⁹There are two reasons for comparing the self-referential group with the total sample instead of comparing it with the remainder group. First, the self-referential group is rather small (n=93), so there are no more statistically significant differences between the self-referential group and the remainder than between the self-referential group and the total sample. Second, the remainder group cannot be defined substantially and is heterogenous because it not only consists of allo-referentials but also of more or less self-referential journalists.

²⁰The cluster analysis used here is neither an hierarchical procedure nor the k-means algorithm provided by SPSS, but a stand-alone program developed by the German sociologist Hans Bardeleben, Conclus 3.0. Professional Cluster Analysis, 2 Volumes (Gießen: Eigenverlag, 1995). It is based on an iterative minimal distance algorithm. In the first step, the respondents are randomly sorted into a given number of clusters. Then, the algorithm exchanges the cases iteratively to optimize the relationship of between variances, which must tend to a maximum, and within variances, which must be minimized. Contrary to the SPSS routines, this program is able to process confirmatory cluster analyis with given centroids.

²¹Armin Scholl and Siegfried Weischenberg, Journalismus in der Gesellschaft. Theorie, Methodologie und Empirie (trans.: Journalism in Society: Theory, Methodology, and Empirical Results) (Opladen, Wiesbaden: Westdeutscher Verlag, 1998), 167 ff. Siegfried Weischenberg, Martin Löffelholz and Armin Scholl, "Journalism in Germany," 242 ff.

²²Siegfried Weischenberg, Martin Löffelholz and Armin Scholl, "Journalism in Germany," 245.

²³Siegfried Weischenberg, Martin Löffelholz and Armin Scholl, "Journalism in Germany," 247 f. Armin Scholl and Siegfried Weischenberg, Journalismus in der Gesellschaft. Theorie, Methodologie und Empirie, 190 f.

²⁴Armin Scholl and Siegfried Weischenberg, Journalismus in der Gesellschaft. Theorie, Methodologie und Empirie, 230.

²⁵We did not find significant differences for all items we asked respondents. Especially in the case of unscrupulous methods this is due to a bottom effect because of low overall agreement with these items, which have only little variances.