Content analysis

Content analysis or textual analysis is a methodology in the social sciences for studying the content of communication. Earl Babbie defines it as "the study of recorded human communications, such as books, websites, paintings and laws."

According to Dr. Farooq Joubish, content analysis is considered a scholarly methodology in the humanities by which texts are studied as to authorship, authenticity, or meaning. This latter subject include philology, hermeneutics, and semiotics.

Harold Lasswell formulated the core questions of content analysis: "Who says what, to whom, why, to what extent and with what effect?." Ole Holsti (1969) offers a broad definition of content analysis as "any technique for making inferences by objectively and systematically identifying specified characteristics of messages." Kimberly A. Neuendorf (2002, p. 10) offers a six-part definition of content analysis:

"Content analysis is a summarising, quantitative analysis of messages that relies on the scientific method (including attention to objectivity, intersubjectivity, a priori design, reliability, validity, generalisability, replicability, and hypothesis testing) and is not limited as to the types of variables that may be measured or the context in which the messages are created or presented."

Description

In 1931, Alfred R Lindesmith developed a methodology to refute existing hypotheses, which became known as a content analysis technique, and it gained popularity in the 1960s by Glaser and is referred to as “The Constant Comparative Method of Qualitative Analysis” in an article published in 1964-65. Glaser and Strauss (1967) referred to their adaptation of it as “Grounded Theory." The method of content analysis enables the researcher to include large amounts of textual information and systematically identify its properties, e.g. the frequencies of most used keywords (KWIC meaning "Key Word in Context") by locating the more important structures of its communication content. Yet such amounts of textual information must be categorised analysis, providing at the end a meaningful reading of content under scrutiny. David Robertson (1976:73-75) for example created a coding frame for a comparison of modes of party competition between British and American parties. It was developed further in 1979 by the Manifesto Research Group aiming at a comparative content-analytic approach on the policy positions of political parties.

Since the 1980s, content analysis has become an increasingly important tool in the measurement of success in public relations (notably media relations) programs and the assessment of media profiles. In these circumstances, content analysis is an element of media evaluation or media analysis. In analyses of this type, data from content analysis is usually combined with media data (circulation, readership, number of viewers and listeners, frequency of publication). It has also been used by futurists to identify trends. In 1982, John Naisbitt published his popular Megatrends, based on content analysis in the US media.

The creation of coding frames is intrinsically related to a creative approach to variables that exert an influence over textual content. In political analysis, these variables could be political scandals, the impact of public opinion polls, sudden events in external politics, inflation etc. Mimetic Convergence, created by F. Lampreia Carvalho for the comparative analysis of electoral proclamations on free-to-air television is an example of creative articulation of variables in content analysis. The methodology describes the construction of party identities during long-term party competitions on TV, from a dynamic perspective, governed by the logic of the contingent. This method aims to capture the contingent logic observed in electoral campaigns by focusing on the repetition and innovation of themes sustained in party broadcasts. According to such post-structuralist perspective from which electoral competition is analysed, the party identities, 'the real' cannot speak without mediations because there is not a natural centre fixing the meaning of a party structure, it rather depends on ad-hoc articulations. There is no empirical reality outside articulations of meaning. Reality is an outcome of power struggles that unify ideas of social structure as a result of contingent interventions. In Brazil, these contingent interventions have proven to be mimetic and convergent rather than divergent and polarised, being integral to the repetition of dichotomised worldviews.

Mimetic Convergence thus aims to show the process of fixation of meaning through discursive articulations that repeat, alter and subvert political issues that come into play. For this reason, parties are not taken as the pure expression of conflicts for the representation of interests (of different classes, religions, ethnic groups (see: Lipset & Rokkan 1967, Lijphart 1984) but attempts to recompose and re-articulate ideas of an absent totality around signifiers gaining positivity.

Every content analysis should depart from a hypothesis. The hypothesis of Mimetic Convergence supports the Downsian interpretation that in general, rational voters converge in the direction of uniform positions in most thematic dimensions. The hypothesis guiding the analysis of Mimetic Convergence between political parties' broadcasts is: 'public opinion polls on vote intention, published throughout campaigns on TV will contribute to successive revisions of candidates' discourses. Candidates re-orient their arguments and thematic selections in part by the signals sent by voters. One must also consider the interference of other kinds of input on electoral propaganda such as internal and external political crises and the arbitrary interference of private interests on the dispute. Moments of internal crisis in disputes between candidates might result from the exhaustion of a certain strategy. The moments of exhaustion might consequently precipitate an inversion in the thematic flux.

As an evaluation approach, content analysis is considered by some to be quasi-evaluation because content analysis judgments need not be based on value statements if the research objective is aimed at presenting subjective experiences. Thus, they can be based on knowledge of everyday lived experiences. Such content analyses are not evaluations. On the other hand, when content analysis judgments are based on values, such studies are evaluations (Frisbie, 1986).

As demonstrated above, only a good scientific hypothesis can lead to the development of a methodology that will allow the empirical description, be it dynamic or static.

Content analysis. This is a closely related if not overlapping kind, often included under the general rubric of “qualitative analysis,” and used primarily in the social sciences. It is “a systematic, replicable technique for compressing many words of text into fewer content categories based on explicit rules of coding” (Stemler 2001). It often involves building and applying a “concept dictionary” or fixed vocabulary of terms on the basis of which words are extracted from the textual data for concording or statistical computation.

Uses of content analysis

Ole Holsti (1969) groups 15 uses of content analysis into three basic categories:

• make inferences about the antecedents of a communication
• describe and make inferences about characteristics of a communication
• make inferences about the effects of a communication.

He also places these uses into the context of the basic communication paradigm.

The following table shows fifteen uses of content analysis in terms of their general purpose, element of the communication paradigm to which they apply, and the general question they are intended to answer.

Uses of Content Analysis by Purpose, Communication Element, and Question
Purpose	Element	Question	Use
Make inferences about the antecedents of communications	Source	Who?	Answer questions of disputed authorship (authorship analysis)
	Encoding process	Why?	Secure political & military intelligence Analyse traits of individuals Infer cultural aspects & change Provide legal & evaluative evidence
Describe & make inferences about the characteristics of communications	Channel	How?	Analyse techniques of persuasion Analyse style
	Message	What?	Describe trends in communication content Relate known characteristics of sources to messages they produce Compare communication content to standards
	Recipient	To whom?	Relate known characteristics of audiences to messages produced for them Describe patterns of communication
Make inferences about the consequences of communications	Decoding process	With what effect?	Measure readability Analyse the flow of information Assess responses to communications
Note. Purpose, communication element, & question from Holsti (1969). Uses primarily from Berelson (1952) as adapted by Holsti (1969).

The process of a content analysis

According to Dr. Klaus Krippendorff (1980 and 2004), six questions must be addressed in every content analysis:

1. Which data are analysed?
2. How are they defined?
3. What is the population from which they are drawn?
4. What is the context relative to which the data are analysed?
5. What are the boundaries of the analysis?
6. What is the target of the inferences?

The assumption is that words and phrases mentioned most often are those reflecting important concerns in every communication. Therefore, quantitative content analysis starts with word frequencies, space measurements (column centimeters/inches in the case of newspapers), time counts (for radio and television time) and keyword frequencies. However, content analysis extends far beyond plain word counts, e.g. with Keyword In Context routines words can be analysed in their specific context to be disambiguated. Synonyms and homonyms can be isolated in accordance to linguistic properties of a language.

Qualitatively, content analysis can involve any kind of analysis where communication content (speech, written text, interviews, images ...) is categorised and classified. In its beginnings, using the first newspapers at the end of 19th century, analysis was done manually by measuring the number of lines and amount of space given a subject. With the rise of common computing facilities like PCs, computer-based methods of analysis are growing in popularity. Answers to open ended questions, newspaper articles, political party manifestoes, medical records or systematic observations in experiments can all be subject to systematic analysis of textual data. By having contents of communication available in form of machine readable texts, the input is analysed for frequencies and coded into categories for building up inferences. Robert Philip Weber (1990) notes: "To make valid inferences from the text, it is important that the classification procedure be reliable in the sense of being consistent: Different people should code the same text in the same way" (p. 12). The validity, inter-coder reliability and intra-coder reliability are subject to intense methodological research efforts over long years (see Krippendorff, 2004).

One more distinction is between the manifest contents (of communication) and its latent meaning. "Manifest" describes what (an author or speaker) definitely has written, while latent meaning describes what an author intended to say/write. Normally, content analysis can only be applied on manifest content; that is, the words, sentences, or texts themselves, rather than their meanings.

Dermot McKeone (1995) has highlighted the difference between prescriptive analysis and open analysis. In prescriptive analysis, the context is a closely defined set of communication parameters (e.g. specific messages, subject matter); open analysis identifies the dominant messages and subject matter within the text.

A further step in analysis is the distinction between dictionary-based (quantitative) approaches and qualitative approaches. Dictionary-based approaches set up a list of categories derived from the frequency list of words and control the distribution of words and their respective categories over the texts. While methods in quantitative content analysis in this way transform observations of found categories into quantitative statistical data, the qualitative content analysis focuses more on the intentionality and its implications.

Reliability in content analysis

Dr. Kimberly A. Neuendorf (2002) suggests that when human coders are used in content analysis, reliability translates to intercoder reliability or "the amount of agreement or correspondence among two or more coders."

See also

• Hermeneutics
• Donald Wayne Foster
• Transition words

References

• Earl Babbie, 'The Practice of Social Research', 10th edition, Wadsworth, Thomson Learning Inc., ISBN 0-534-62029-9

• Bernard Berelson: Content Analysis in Communication Research. Glencoe, Ill: Free Press 1971 (first edition from 1952)

• Ian Budge, Hans-Dieter Klingemann et al.: Mapping Policy Preferences. Estimates for Parties, Electors and Governments 1945-1998. Oxford 2001: Oxford University Press, ISBN 0-19-924400-6 (great example of application of content analysis methods in Political Science dealing with political parties and its impact on electoral systems)

• Richard Frisbie: The use of microcomputer programs to improve the reliability and validity of content analysis in evaluation. Paper presented at the session, Qualitative research methods, at the Annual Meeting of the American Educational Research Association, San Francisco. 1986, April

• Graneheim, Ulla Hällgren, & Lundman, Berit (2004). Qualitative content analysis in nursing research: concepts, procedures and measures to achieve trustworthiness. Nurse Education Today, 24(2), 105-112.

• Ole R. Holsti: Content Analysis for the Social Sciences and Humanities. Reading, MA: Addison-Wesley. 1969

• Lampreia Carvalho, F. 'Continuity and Innovation: Conservatism and Politics of Communication in Brazil'. (Continuidade e Inovacao: Conservadorismo e Politica da Comunicacao no Brasil) Journal Revista Brasileira de Ciencias Sociais, N.43, 2000. ANPOCS, Brazil.

• Klaus Krippendorff: Content Analysis: An Introduction to Its Methodology. 2nd edition, Thousand Oaks, CA: Sage 2004

• Klaus Krippendorff and Mary Angela Bock (editors): The Content Analysis Reader. Thousand Oaks, CA: Sage Publications. 2008

• Dermot McKeone: Measuring Your Media Profile, Gower Press, 1995 A general introduction to media analysis and PR evaluation for the communications industry

• Neuendorf, Kimberly A.: The Content Analysis Guidebook Thousand Oaks, CA: Sage Publications. 2002
• Neuendorf, Kimberly A.: The Content Analysis Guidebook Online (2002): http://academic.csuohio.edu/kneuendorf/content/

• Carl W. Roberts (ed.): Text Analysis for the Social Sciences: Methods for Drawing Inferences from Texts and Transcripts. Mahwah, NJ: Lawrence Erlbaum 1997

• Robert Philip Weber: Basic Content Analysis. 2nd ed., Newbury Park, CA: Sage 1990 (recommended introductory reading)

• Roger D. Wimmer and Joseph R. Dominick Mass Media Research: An Introduction. 8th ed. (Belmont, CA: Wadsworth, 2005).

• Stemler Steve: An Overview of Content Analysis (2001):http://pareonline.net/getvn.asp?v=7&n=17

External links

• http://www.yoshikoder.org/downloads.html: an open source content analysis program
• http://academic.csuohio.edu/kneuendorf/content/ Web site of the Content Analysis Guidebook Online, provides some CATA software for free download, list of archives, bibliographies and other important sources
• http://www.impact-media.co.uk Contains a general introduction to media analysis and media profile measurement including an outline of the differences between open and prescriptive analysis
• http://www.andersonanalytics.com/reports/AATAT.pdf: History of content analysis software in psychology and applications of content analysis, text mining and text analytics in market research.
• http://courses.washington.edu/socw580/contentsoftware.shtml - Software for Content Analysis, a list of programs for analyzing text, images, video and audio; a resource used by the Advanced Research Methods & Design course in the School of Social Work at the University of Washington.