Kongruenz der Anlassauswahl als Indikator für die Journalismusforschung: Eine Exploration

Mit einer Inhaltsanalyse untersucht der Beitrag ressortvergleichend: 1. wie häufig dieselben nachrichtlichen Anlässe von mehreren Medientiteln zur Berichterstattung ausgewählt werden und 2. ob diesen (in-)kongruenten Auswahlentscheidungen eine Gesetzmäßigkeit zugrunde liegt. Das Interesse gilt dabei besonders der Kongruenz der Nachrichtenauswahl in Wissenschaftsressorts. Argumentiert wird, dass die Verteilung der Auswahlkongruenz Rückschlüsse auf ressortspezifisch unterschiedliche Bedingungen der Nachrichtengebung erlaubt. Diese wirken sich auf die Leistungsfähigkeit des Journalismus aus, die öffentliche Aufmerksamkeit auf besonders relevante Ereignisse zu lenken.
... mehr

Analysiert wurden gut 4000 Artikel, die 2018 und 2019 in einem Zeitraum von einer bzw. zwei Wochen in fünf deutschsprachigen Zeitungen erschienen sind und 2521 verschiedene Anlässe hatten. Der Anteil der Anlässe, die kongruent von mehreren Zeitungen zugleich ausgewählt wurden, nimmt mit dem Grad der Kongruenz exponentiell ab, was frühere Ergebnisse bestätigt. Im Vergleich der verschiedenen Ressorts ist der Anteil exklusiver Anlässe im Wissenschaftsressort deutlich höher. Wissenschaftsredaktionen wählen wesentlich inkongruenter aus als beispielsweise Politikredaktionen. Diese Ergebnisse wecken Zweifel daran, ob es speziell dem Wissenschaftsjournalismus gelingt, das öffentliche Interesse an einem Kernbestand von wichtigen Ereignissen zu befördern.

Each day, journalists have to select from a large number of statements, social and natural events those that they believe will attract the interest of recipients and thus generate attention. These selections are influenced by a number of factual, temporal and social constraints that differ between journalistic fields. We aim to identify and at least partially explain an “external effect” of these selection decisions, a social phenomenon resulting from the aggregation of the actions of individual journalists. The effect we are interested in is the congruence of the event selection, i.e. ... mehrthe number of specific events that are selected congruently by the departments of two, three, four, etc. different media titles to initiate reporting. Our focus is especially on the selection of events in science departments.

The analysis of this external effect is relevant for two reasons: firstly, the congruent, focussed selection of events or topics is an important factor in a number of central concepts in mass communication research. For example, a minimum of congruent news coverage is a basic requirement for agenda-setting effects. It is also part of two conflicting normative demands on journalism: less congruence fosters the diversity of journalistic coverage while more congruence is necessary for a focus on individual, particularly important events. Secondly, we argue that the distribution of congruent choices can be used as an indicator for differences in the conditions of news production in different news departments.

Prior research indicates that a congruent selection of events by science departments is rather rare. In an earlier work, we conducted an input-output-analysis of the coverage of study results published between August 2014 and July 2018. In this study, only eight of 10,000 study results (0.08%) were taken up congruently by journalists worldwide. Additionally, we found that the share of studies selected by a certain number of media outlets decreased regularly with an increasing number of media outlets. While 87 studies were selected by 55 outlets, eight studies were selected by 100 outlets and only one by 200 outlets. Thus, most studies receive a small amount of journalistic attention (if they are selected at all), but some results are selected by a disproportionate number of journalists, leading to a distribution that follows a function of the form 1/n3.4.

Similar distributions were found for the selection of events by political journalists in multiple studies conducted by Patrick Rössler. Compared with the distribution of the selection of study results, a larger share of events was selected by many or all media titles in his samples, leading to a more congruent selection. Based on these results, we pursue two goals: we try to reproduce this distribution type in a systematic sample of media titles and we examine whether the editorial selection of events is more incongruent in science departments than in other departments.

To achieve both goals, we conducted a quantitative content analysis in which we determined how many coverage occasions were selected congruently by a small sample of five media outlets. We have analysed more than 4000 articles that appeared in the German daily newspapers Die Welt, Der Tagesspiegel, Süddeutsche Zeitung, Frankfurter Allgemeine Zeitung and Stuttgarter Zeitung over periods of one respectively two weeks in 2018 and 2019. These newspapers were chosen because they are rather similar in their circulation and profile and especially because they all have science departments responsible for a daily or weekly section of the newspaper. All articles published in the science section, the title page and politics section and, only for the first investigation period, all articles in the economics, culture, media and miscellaneous (Vermischtes) sections were included in the analysis. We recorded the topic of each article as well as the type of event that initiated the article. The concrete event was coded as an open text variable. The list of concrete events assembled during the coding was used to identify the share of exclusively or congruently selected occasions.

In total, we found 2521 different events leading to the 4112 articles in our dataset. The 279 articles published in the science sections of the five newspapers were triggered by 252 different events, leading to a share of 92.1% of all events that was selected by only one media outlet. No event was selected by all five or even four newspapers. With increasing congruence, the number of events selected by a particular number of newspapers decreases by 1/n3,3. The distribution curve thus decreases by almost exactly the same factor as the congruence of the selection of study results, which can be interpreted as a strong indication that these distributions are scale invariant. In comparison with the science sections, news selection was much more focussed in the politics section of the newspapers. Of 970 events found in this section, 72% were selected exclusively, 27 events (or 2.8%) were selected by all five newspapers. The share of events decreases regularly with increasing congruence as well, this time by a much smaller factor of 1/n2,1. In the other sections, we find similar distributions with exponents between those found in the science and politics sections.

The comparison with the results from our earlier work shows that the exponent of the congruent selection decisions in the science departments of five media titles corresponds almost exactly to that determined for several thousand online media titles. This gives us reason to assume that the distribution of the social dissemination of individual events via German journalistic mass media with independently selecting science departments can be generalized as follows: The probability that the same event will be taken up congruently decreases by a factor of approximately 1/n3,3 irrespective of the number of media titles. Compared especially with political journalism, science departments show a rather incongruent selection of events. These results cast doubt on the ability of science journalism to focus public attention on single relevant scientific events.