Content analysis

from Wikipedia, the free encyclopedia

The content analysis ( English content analysis ) is a method bundle of empirical social sciences. The object is the analysis of the content of communication , which is available in the form of texts, e.g. as written interviews or newspaper articles.

Content analysis techniques are generally understood to mean the quantitative evaluation of texts that can be used alone or in conjunction with qualitative methods such as qualitative content analysis or discourse analysis .

Basics

Bernard Berelson and Harold D. Lasswell with studies of war reports and war propaganda during the Second World War are considered the founders of content analysis . Early civil applications of content analysis can be seen in the early years of the 20th century, while important developments date back to after the end of World War II. One of the earliest works of this kind is likely to be that of the Hamburg doctor Ernst Rittershaus. Under the title "Irrsinn und Presse. Ein Kulturbild" (Insanity and Press. A Culture Picture) in 1913, on 245 pages, he explained in detail, if not systematically, how seven Hamburg newspapers reported on topics such as psychiatry , mental illness , insane asylums, alcoholism and crime .

The material for a content analysis according to today's requirements is coded and disclosed according to a (scientifically) justifiable scheme . In the quantitative content analysis, the researcher proceeds in a theory- guided manner, i.e. not as in the hermeneutic text interpretation , in which the meaning (content) of the text and the intention of the author are the subject of the analysis, but interesting aspects are isolated and based on (social) scientific hypotheses evaluated with the help of empirical methods of content analysis.

The aim of content analyzes is to reduce the existing complexity of the research material (the texts). This is achieved by hiding the uninteresting part of the communication using stop word lists and, after a frequency analysis, the keywords are subjected to an analysis of the “Keywords in context” KWIC . KWIC analyzes are fundamental first steps on the way to a content analysis, but not the final one. This is followed by further steps that allow a quantitative-statistical evaluation of communication content.

The formal-descriptive approach examines external text features. The diagnostic approach tries to make statements about the conditions in which the material was created or about the motives for reporting (first meaning of inference ). The prognostic approach tries to make a prognosis about the effect of the content on the recipient (second meaning of inference).

A distinction is made between different working methods of content analysis: Frequency analyzes are limited to a descriptive counting of the word frequencies. In Valenzanalysen is additionally detected whether the contents are evaluated positively or negatively. The intensity analysis ascertains the intensity of assessments with the help of suitable scales .

A number of computer programs exist to support content analysis methods. By facilitating the availability of computing capacity in recent times computerized important increase in the quantitative content analysis, the (so-called. The original text using word lists dictionary-based approach ), the structure of the text in a statistically be evaluated format transform (see FIG. Statistics ). In content analysis, statistical methods that are traditionally attributed to "exploratory data analysis" are particularly suitable, such as multidimensional scaling , the similarities between the individual texts, their speakers or authors and the underlying concepts trying to measure and model. Further statistical possibilities correspondence analysis, cluster analysis and multi-level analysis (Engl. Multi-level analysis ).

Content analysis process

In the planning phase , the research question is first formulated and suitable hypotheses are drawn up, which are often derived from more comprehensive social science theories (such as: theories of organization, theories of political decision-making, etc.). Then, on the basis of the research question and the hypotheses, the research material is selected ( research units, e.g. issues of certain newspapers in a certain period of time, television programs of certain stations to be investigated). If a full survey is not possible, not useful or too expensive, a sample is drawn from the material collected.

In the development phase is with theory-based and empirically guided education category , a category system created (dictionary-based approach), the basis of which an investigation is conducted. The category system should be comprehensive, i.e. it should cover all aspects of the research material that must be taken into account in order to answer the research question. In addition, it must be disjoint and unambiguous, i.e. each feature of the examined content should only be able to be assigned to one category without any doubt. A category system that is too detailed should be avoided because the uniqueness of the system suffers from overly differentiated categories and practical handling is made more difficult. Coding rules are then formulated. These are specific instructions to human coders as to how the examined content is encrypted, i.e. coded in numbers , or the specific criteria according to which a computer-aided content analysis is carried out. Category scheme, coding instructions and coding sheet (this is a data entry mask in printed or electronic form in which the codes for the analyzed content are entered) are later summarized in the code book .

In the test phase , a test coding ( pretest ) is carried out and, if necessary, the category system is revised to ensure the highest possible reliability . The intracoder reliability (the same coding result with an identical encoder in different coding processes) and the intercoder reliability (the same coding result from different coders with identical material) are empirically checked.

When the tests are done, the actual coding takes place. To prepare for the coding phase , human coders have to be trained extensively and the machines have to be programmed for computer-aided content analysis. The test material should be distributed as randomly as possible to the coders so that unavoidable errors and possible peculiarities of the coders when using the codebook are randomly distributed to the codings.

In the evaluation phase, the data obtained from the texts are processed, checked and evaluated using statistical methods, such as multidimensional scaling (MDS), variance analysis (ANOVA, MANOVA ), correspondence analysis or network analyzes . In line with the scaling level of texts, these methods have recently increased in importance, while factor analyzes and cluster analyzes have tended to decrease in number and importance.

Evaluation and assessment of the method

Content analysis does not necessarily depend on cooperation with respondents (problem of willingness to participate or reactivity ). Because texts can be compiled relatively easily thanks to modern means of communication (websites, archives and libraries for written material ...), it is a preferred application for secondary analyzes. Similar to survey research, the basic problem remains, to what extent the selection of the texts is later statistical representative statement (about a group of actors, the population) allowed. Content analysis alone cannot make statements about effects or about intentions (of the author, speaker). This requires a supplementary communicator or recipient survey (communication research). The descriptive data serve as a prerequisite for subsequent impact analyzes. Historically, the content analysis has repeatedly shown that it is suitable as an early warning system (used in the analysis of German war propaganda in World War II). As a method of data evaluation, content analysis offers interesting aspects where non-numerical raw material (i.e. texts) is transformed into a mathematically-statistically accessible form that can be analyzed using a computer. It is essential to observe that all communication takes place in one or more contexts at the same time: the relation of communication to the recipient (recipient), intentionality, but also extra-linguistic elements (body language) can play both reinforcing and contradicting roles for communication. Compared to other methods, content analysis has the following advantages:

  • With the help of content analysis, statements can be made about communicators and their intentions that are not or can no longer be reached. Through systematic (computer-aided) analysis, properties can be mapped that suggest a similarity or dissimilarity of texts or the underlying concepts of the respective authors / speakers (e.g. with the help of multidimensional scaling).
  • The examination material is available regardless of time.
  • There is no risk of reactivity of the test material.
  • The examination can be carried out repeatedly (replication, secondary analysis) and the procedure can be followed.
  • It is comparatively inexpensive.

Applications for content analysis can be found in the whole of the social sciences: in political sociology in examining election programs and political texts (parliamentary debates), in psychology in examining diary entries, in education in examining student essays, in linguistics and criminology when clarifying the authorship of anonymous texts using sample comparisons, in marketing when examining advertising texts or statements, in consumer research (popularity of products) and when comparing legal texts.

See also

literature

Theoretical representations

  • Robert Philip Weber: Basic Content Analysis (=  Quantitative Applications in the Social Sciences . Volume 49 ). 2nd Edition. Sage Publications , August 1, 1990 (English).
  • Werner Früh: Content analysis: theory and practice. 6., revised. Edition. UVK, Konstanz 2007, ISBN 978-3-8252-2501-8 .
  • Patrick Rössler: content analysis. UVK, Konstanz 2005. (a current textbook on the standardized content analysis described here)
  • Klaus Krippendorff: Content Analysis: An Introduction to Its Methodology. 2nd Edition. Sage, Thousand Oaks (CA) 2004. (English-language basic literature. Already the first edition. From 1980 was considered a classic that presented the entirety of the content analysis method).
  • Peter Atteslander : Methods of Empirical Social Research. 10th edition. de Gruyter, Berlin 2003, pp. 215–249 (the chapter on content analysis also includes computer-aided content analysis, which is discussed briefly in Früh. Methods of qualitative analysis methods are also dealt with)
  • Ian Budge, Hans-Dieter Klingemann and others: Mapping Policy Preferences. Estimates for Parties, Electors and Governments 1945–1998. Oxford University Press, Oxford 2001, ISBN 0-19-924400-6 . (an English-language presentation and exemplary application of content analysis in the field of political science )
  • Klaus Merten: Content analysis: introduction to theory, method and practice. 2nd, improved edition. Westdeutscher Verlag, Opladen 1995 (this book provides a good breakdown of the various methods, summarized under the term content analysis. It also refers to further literature.)
  • Ekkehard Mochmann: content analysis. In: Jürgen Kriz , D. Nohlen , RO Schultze (Ed.): Lexicon of Politics. Volume 2: Political Science Methods. Verlag CH Beck, Munich 1994, ISBN 3-406-36904-9 , pp. 184-187. (Brief presentation with an outline of the historical development of content analysis as a method, new edition published in 2002. Highly recommended as an introductory text)
  • Cornelia Züll, Peter P. Mohler (Ed.): Text analysis. Applications of computer-aided content analysis. Westdeutscher Verlag, Opladen 1992. (especially for social science users)
  • Ralf Lisch , Jürgen Kriz: Basics and models of content analysis. Inventory and criticism. Reinbek near Hamburg: Rowohlt Taschenbuch Verlag, 1978, ISBN 3-499-21117-3 (comprehensive presentation and critical discussion of different content analysis approaches and their methodological bases)
  • Alphons Silbermann : Systematic content analysis. In: René König (Ed.): Handbook of empirical social research. Volume 4, Stuttgart 1974, pp. 253-339. (basic German language representation)
  • Jürgen Ritsert : content analysis and ideology criticism. An essay on critical social research. Frankfurt am Main 1972. (for an overview of the older critical discussion of quantitative content analysis)

Applied content analysis

  • Michael Bröning: "Like any other country ..." The image of Germany in the Israeli press 1990–2000 . Münster: LIT 2004. ISBN 3-8258-7452-4 . (The book also provides impulses for the methodological development of content analysis; at the same time Diss. Kassel 2003)
  • Jürgen Gerhards , Mike S. Schäfer: The creation of a public hegemony: Human genome research in the German and US American press . Wiesbaden: Verlag für Sozialwissenschaften 2006, ISBN 978-3-531-14964-6 .
  • Werner Gruhn: Science and technology in German mass media. A comparison between the Federal Republic of Germany and the GDR . Erlangen: German Society for Contemporary History 1979, ISBN 3-88150-042-1 . (With explanations on content analysis in two different political and media systems)
  • Constanze Herweg: Medicine in words and pictures in the GEO reportage magazine . Berlin: LIT 2011, ISBN 978-3-643-10331-4 . (also phil. diss. Bochum 2005)
  • Ulrike Hoffmann-Richter : Psychiatry in the newspaper. Judgments and prejudices . Bonn: Edition Das Narrenschiff in Psychiatrie-Verlag 2000, ISBN 3-88414-295-X
  • Horst Merscheim: Medicine on TV. Problem of mass media mediated health reporting . Bochum: Studienverlag Dr. N. Brockmeyer 1984, ISBN 3-88339-358-4 . (At the same time phil. Diss. Bochum 1983)
  • Kerstin Reinisch: Choice without knowledge? Empirical analysis on the depoliticization of election campaign reporting in German print media. Baden-Baden: Nomos 2017, ISBN 978-3-8487-4060-4 .
  • E (rnst) Rittershaus: Insanity and the press. A cultural image . Jena: Gustav Fischer 1913 (very early press analysis, still without systematics and coding)
  • Eckart Roloff : The journalistic discovery of the patient. A press analysis on medical journalism and the first heart transplants . Baden-Baden: Nomos 2013, ISBN 978-3-8487-0731-7 . (Slightly edited version of a Salzburg dissertation from 1972 with a detailed foreword and epilogue from 2013, detailed description of the content analysis system)
  • Mike S. Schäfer : Science in the media. The medialization of scientific topics . Wiesbaden: Verlag für Sozialwissenschaften 2007, ISBN 978-3-531-15592-0 .
  • Martin Sturmer: Africa! Plea for differentiated reporting . Constance: UVK 2013, ISBN 978-3-86764-323-8 . (Example of the combination of content and media response analysis)

Web links

Wiktionary: content analysis  - explanations of meanings, word origins, synonyms, translations

Individual evidence

  1. ↑ Insanity and the press: Ein Kulturbild, Fischer, Jena 1913.