Content and Discourse Analysis

Leonie Newnham, Jean-Jacques Pantebre, Mike Spark

Content analysis

Content analysis is a sub-set of secondary data analysis (i.e., analysis performed on data collected by others), and has been noted in Websters Dictionary since 1945. It is the analysis of the manifest and latent content of a body of communicated material (as a book or film) through a classification, tabulation and evaluation of its key symbols and themes in order to ascertain its meaning and probable effect (Webster 1).

Secondary data can encompass census data as well as unobtrusive measures , such as physical traces, simple observation and archival records Ð it is the last that are the major subject of content analysis. It is a research technique for the objective, quantitative and systematic study of communication content. It involves charting or counting the incidence, or co-incidence, or particular items belonging to a set of (usually) predetermined categories. (Jary & Jary, p. 119)

Researchers specify the content characteristics to be analyzed and apply rules for identifying and recording them as they appear in the archived records and documents. These rules must be explicit and replicable, so that the researcher cannot select only the verbal and written occurrences which support his/her hypothesis.

Themes can be examined as well, and content analysis has also been used to infer elements of culture and cultural change (McClelland). It can be performed informally, such as when one reads an editorial in one newspaper and decides that the paper has changed its stance on a particular issue, based on ones reading of not only what’s there but also what’s not there.

Further reading of other newspapers could have one decide that the climate has changed as all have changed their approach or stance.

Discourse analysis is a slightly more recent (1952) and presumably higher-level effort, being the study of linguistic relations and structures in discourse (Webster 2). It is defined in the Collins Dictionary (Jary & Jary, p.168) as forms of textual analysis in which the aim is to exhibit the structure of discourse and discourse formations" or analysis of the role of the reader or viewer in reading and creating meaning. 'The assumption is that a discourse has identifiable formation rules which distinguish it from other discourses.

Advances in computer technology have greatly expanded the ease of usage and accuracy of content analysis. Only when the software for conducting content analysis on desktop computers becomes as convenient as word-processing or spell-checker programs -and this should only be a matter of time - are many investigators again likely to utilize it in their research , stated Borgatta in 1992 (p. 292). Advances in our knowledge of artificial intelligence during the 1990s are similarly likely to extend the usage of this research tool.
In historical analysis, it is often the only means available to the researcher. However, if we are concerned with analysis of the larger social picture, one has to ensure that the material is not so restrictive, either in points of view or in actual numbers, as to make it unrepresentative.

It is relatively inexpensive research in terms of financial cost, as the data is normally found in public or accessible private archives. Much national and even international research can be done with little funds, from the comfort of ones own location. The increasing access to the Internet will also greatly assist the adoption of this form of research in many areas. However, the main research cost will continue to be in the large amount of time expended on setting up the study parameters, and the actual data logging and collection.
In preparing the study specifications, the sublimation of the researchers personal attitudes, biases and assumptions must be complete, or the outcomes or conclusions derived from the study can quite rightly be called into question.

As with other secondary data analysis forms, a major disadvantage is that the researcher has no control over the original data collection and preliminary analysis - if one is making conclusions from the data, one must ascertain the accuracy, validity and time-frame of the data, and deduce with caution.

References

Borgatta, E. & Borgatta M., Encyclopedia of Sociology, Vol 1, (1992), New York, MacMillan

Jary, D. & Jary, J. Collins Dictionary of Sociology, 1995, Glasgow, Harper Collins.

McClelland, David C. The use of Measures of Human motivation in the Study of
society in Motives in Fantasy, Action and society, ed. John W. Atkinson (New York:
Van Nostrand, 1966), p.518.

Webster 1& 2 - wysiwyg://23/http://www.m-w.com/cgi-bin/netdict

Content Analysis

Fishe, J., 1982. Introduction to Communication Studies. New York: Methuen & Co.

Fishe (1982) provides a succinct definition of content analysis. It is designed to produce an objective, measurable, verifiable account of the manifest content of messages (p.119). It analyzes the denotative order of signification and works best on a large scale as the more it deals with the more accurate it becomes.

Applying it to its use in the study of communication, Fishe notes that content analysis works through identifying and counting chosen units in a communication system. The units are open to choice by the researcher and only need to be readily identifiable and occur frequently enough for statistical methods of analysis to be valid. It is non selective and must cover the whole message system or a properly constituted sample.

In the section of this book on content analysis, a number of case studies of research undertaken using this method were reviewed as a way of explaining how it can and has been used. For example as ways of checking the subjective and selective way we normally receive messages through various media such as television, a study of job stereotyping on American fictional television, Seggar and Wheeler (1973) found that there was a far more restricted range of occupations for women compared to men.

Form can be studied as well as content when reviewing some media such as television commercials for how actively boys were portrayed as opposed to girls. Content analysis can reveal a media distortion in terms of a particular message however it cannot identify why that exists. It can provide data upon which to base discussions of larger social issues that could be the basis for the phenomenon observed.

George Gerbner is identified as having produced the most fully developed theory of how content analysis can shed light on deeper cultural matters. He believes that culture communicates with itself through the mass media and reinforces the broad consensus of values in a culture.

Content analysis by having the ability to analyze the whole message system is able to overcome the selectiveness of the individual. Content analysis can reveal patterns and frequencies within the denotative order of communication that connate values and attitudes. Fishe notes that some general laws can be related to content analysis in the denotative order to connotations of social values. For example being a victim on television is a metaphor for being of low status in real life.

Fishe’s book reviews content analysis as it applies to communication studies. It reveals the ability of this form of research to reveal larger cultural issues when used in this field. While it is acknowledged there is a difficulty in using information from this research to identify why certain messages are manifested, it can indicate connotations of social values. This differs from its use in other disciplines where there appears to be caution with the use of the research outcomes from content analysis to explain what they indicate in the larger picture.

Kolbe,R.H. and, Burnett, M.S., 1991. Content-analysis research: an examination of applications with directives for improving research reliability and objectivity. Journal of Consumer Research, Sept. 1991 Vol.18 No2. Pages 243- 251.

This article reviews content analysis as used in consumer communications. It studies a number of research pieces completed to review how research was meeting research methodological standards. Through this a number of benchmarks established through the work of Kassarjian (1977) on the use of content analysis in consumer research. The work identified that content analysis offers the benefits to consumer researchers of allowing for an unobtrusive appraisal of communications and can assess the effects of environmental variables that can be regulatory, economic and cultural. Source characteristics such as attractiveness, credibility and likeability can also be related to message content. Different kinds of message content can impact on receiver responses and have cognitive, affective and behavioural effects. This type of research can also provide a starting point for new research evidence about the nature and effect of specific communications.

The assessment of the critical methodological issues identified that there was difficulty meeting the objectivity and reliability requirements of the research method. The measures used by the authors to review the various components of the research identified a number of categories to measure objectivity, to establish hypothesis testing, review data collection designs and develop reliable categories and coding designs. It was noted by the researchers that content analysis was an important method for facilitating many other types of analyzes for example in protocol analysis, process analysis and integrative literature reviews. The authors identified that it can also assist in theory development and they identified five roles for content analysis in theory development using Lijpharts (1971) categorizations for case study research.

This research method can embellish, augment, accumulate and describe information. The need for systematic study and information acquisition, part of the initial steps in theory development, can be readily provided by content-analysis research.p.99.

Further references

Kassarjian,H.H., 1977. Content Analysis in Consumer Research. Journal of Consumer Research, June Vol.4. Pages 8-18.
Lijphart,A., 1971. Comparative Politics and the Comparative Method. American Political Science Review,Sept. 1971 Vol.65. Pages 682-693.

Minichiello,V., Aroni, R., Timewell,E., and, Alexander,L., 1990. In-depth Interviewing- Researching People. Melbourne: Longman Cheshire. 1990.

This work reviews content analysis as an adjunct to research that incorporates interviewing as a data collection method. As a work that concentrates on the practical how-to of interviewing, content analysis is reviewed as a way to make sense of the information that has been obtained through this process. They identify that this could be used in the perhaps more traditional way of reviewing messages or elements in terms of their frequency of appearance as they appear in interview transcripts or written documents such as diaries, etc. However the authors identify that meaning should be examined as well.

The elements or units of analysis they identify are words, concepts, sentences, and themes. Here they distinguish between the manifest context or elements that are physically present and countable, and the latent content, or the symbolism underlying the physically presented data. The latter is open to interpretation by the researcher and introduces the need to consider how interpretation is consistent with the informants perspective.

With this extension of the idea of content analysis is the need for the interviewer to try to understand the patterns of meaning in the conversation and actively work towards empathic or verstehen understanding. Also the researcher operates within the constraints of the knowledge of the informant and the researcher. Participant concepts, expressed in their everyday language have to be interpreted into theoretical concepts by the researcher for analysis. This broader use of content analysis requires a review of the whole text to extract the essence of the informants meanings.

Logical analysis is a technique that can be used for this process. Techniques for doing qualitative data analysis are outlined. Developing a coding system is reviewed and ways to develop codes and strategies for doing this outlined.

Coding the data is discussed including a brief discussion on index cards and computer packages such as NUDIST. A number of case studies are provided to assist in understanding. This book provides an interesting perspective on content analysis that introduces the use of this with interviews conducted by the researcher as opposed to already published sources. It explores the need to include an interpretative component to assist with coding to ensure that the researcher provides the correct understanding of the piece. This introduces another level of complexity with the use of the data that needs to be considered in the context of the ability of one person operating in one frame of reference to interpret information presented by a person operating in a different frame.

The book expresses the issues without exploring them in great depth theoretically and works through them on a more practical level through the case studies.

Morris,Rebecca., 1994, Computerized Content Analysis in Management Research: a demonstration of advantages and limitations. Journal of Management, Winter 1994, v20 n4 p903(29)

The use of content analysis in management research is reviewed in this article along with the ability to use computerized approaches to this method. Management research has used this technique primarily to draw valid inferences from the textual communications of managers. Content analysis allows the values, sentiments, intentions, and ideologies of managers to be studied in an unobtrusive manner. It also provides a methodology for the systematic analysis of information contained in corporate documents and so opens a rich data source.

The objectives for which content analysis is used in social science are related to the focus of many areas of management research. These being: 1. To make inferences about the values, sentiments, intentions or ideologies of the sources or authors of the communications; 2. To infer group or societal values through the content of communications; 3. To evaluate the effects of communications on the audiences they reach.

Computer content coding schemes are able to formalise coding rules through the creation of computer content-coding schemes, this provides perfect coding reliability is obtained as well as the ability to analyze large volumes of written communication. This has been used in other disciplines such as psychology and sociology.

The author reviewed the use of computerized content analysis in management research by comparing the analysis of human coders to computerized coding of the same text communications. She found that the systems were able to provide a number of advantages including: · Perfect stability of the coding scheme as the computer applies the rules in the same way; · Explicit coding rules yielding formally comparable results; · Perfect coder reliability of the computerized approach allowing the researcher to concentrate on other parts of the study; · Easy manipulation of the text to allow statistical analysis of factors such as word frequency and allow further research of certain terms or themes; · Ability to process larger volumes of qualitative data at lower cost.

There were limitations as the computer programs do not have natural language processing capabilities in the software and so are unable to assess ambiguous concepts and identify the communicative intent of word usage. Programs cannot resolve references back or forwards and it cannot analyze data as strips. Also the researcher is unable to provide an exhaustive list of key words for categories that are by nature indeterminate. The use of computers can provide results that are meaningless as they use literal processing algorithms that result in word crunching without context or meaning. Continued reliance on human coders for the more subtle analysis will be needed.

The author completed the article by an analysis of ways in which Content Analysis could be used in management research.

Neuman, W.L., 1991. Social Research Methods, Qualitative and Quantitative Approaches. Boston: Allyn and Bacon, Boston, 1997.

This is an excellent general research methods book that explains and outlines content research in Chapter 11 titled, Nonreactive Research and Available Data. Neuman identifies nonreactive research as beginning when a researcher notices something that indicates a variable of interest and uses nonreactive or unobtrusive methods. He identifies the key being that the people being studied are not aware of the study and so leave evidence of their social behaviour or actions naturally.

A researcher first conceptualises a construct, then links this to nonreactive empirical evidence which is used as the measure and uses the logic of quantitative measurement for further examination. Neuman identifies content analysis within this framework purely as a technique for gathering and analyzing the context of text. Context can refer to words, meanings, pictures, symbols, ideas, themes, or any message that can be communicated. The text is anything written, visual, or spoken that serves as a medium for communication.

Objective and systematic counting and recording procedures are used to produce a quantitative description of the symbolic context of the text. Neuman emphasises the quantitative rather than the interpretive version of content analysis. He notes that as Markoff, Shapiro and Weitman (1974) suggested it might be better named textural coding. While identifying that content analysis has been used for almost a century and is used in many fields, including literature, history, journalism, political science, education, psychology, it is noted by Neuman that the qualitative content analysis is not highly respected by positivist researchers.

Topics appropriate for content analysis are identified in this book. It identifies the range of studies that have been done including studying themes in popular songs, trends in topics newspapers cover, sex-role stereotypes in text books or films, gender differences in conversations. There is a reference to work by Seider (1974) who content analyzed the public speakers of U.S. corporate executives. This identified five ideological themes that executives emphasised more or less depending on the industry of their corporation.

Woodrum(1984:1) noted that content analysis had wider applicability within social science for studying beliefs, attitudes organizations and human relations.

Neuman identified it being useful for three types of research problems: · those involving a large volume of text; · those that must be studied at a distance; · Those intendedto reveal messages in a text that are difficult to see with casual observation.

Measurement and coding are reviewed. General issues revolve around developing careful measures as often the researcher is dealing with symbolic communication and trying to make it precise, objective and quantitative. Units of analysis are determined by the researcher and can include a range of other units including recording units, context units, and enumeration units. What is measured is based on structured observation based on written rules.

Coding systems are identified that identify characteristics of text content such as frequency of occurrence, direction of message content, intensity of the message and space allocated to the message.

Types of coding include: · manifest coding that records visible, surface content in a text but does not consider the connotations of the word or phrase; · latent coding (also semantic analysis) that looks for underlying, implicit meaning in the content of a text; Intercoder reliability is often an issue where a number of people are working on a large project. This must be catered for in the design of the research and taking care with training about the use of the coding system.

Content analysis research starts with the question formulation, a determination of the units of analysis and an identification of sampling techniques. This is explained in this book by the development of a case study during which a research project is established through the question formulation, to developing variables and coding categories and a process for recording.

Neuman identifies as critical the ability of the researcher to be able to determine what inferences can be made by the researcher about the results of content analysis research. As it describes what is in the text it cannot reveal the intentions of those who created it or the effects that messages in the text have on those who receive them. There is a limit to inferring causality or testing a theory based on nonreactive data accordingly.
This work is a good starting point for understanding content analysis as used in a quantitative manner and using already published material. It does not explain or review using content analysis in combination with other research techniques such as interviews or case studies. By classifying it as nonreactive research, its ability to be used on a wider basis is ignored.

Further Reading

Bouma, G.D.,1996 The Research Process. Melbourne, Oxford University Press,Third Edition:, 1998.
Burns, R.L.,1998 An Introduction to Research Methods. South Melbourne, Longmans,Third Edition:, 1998.
Babbie, E.R., 1991. The Practice of Social Research. Belmont,CA, Eighth Edition: Wadsworth Publishing, Boston, 1998.
Kvale,S., 1991. Interviews. Thousand Oaks,CA:
Weber, R.P.,1985 Basic Content Analysis. Beverly Hills, CA., Sage.

The use of content analysis in business research

Content analysis has been used for almost a century and is used in many fields, including literature, history, journalism, political science, education, psychology. Neuman (1991) identified it being useful for three types of research problems: · Those involving a large volume of text; · Those that must be studied at a distance; To reveal messages in a text that are difficult to see with casual observation.

Within management research it can be used as follows: · To make inferences about the values, sentiments, intentions or ideologies of the sources or authors of the communications; · To infer group or societal values through the content of communications; · To evaluate the effects of communications on the audiences they reach.

Content Analysis is a technique for gathering and analyzing the context of text. Context can refer to words, meanings, pictures, symbols, ideas, themes, or any message that can be communicated. The text is anything written, visual, or spoken that serves as a medium for communication. Objective and systematic counting and recording procedures are used to produce a quantitative description of the symbolic context of the text.

There is a wealth of textural material that can be analyzed from existing sources within organizations including annual reports, speeches given by leaders, and management communications. Material can be assessed over time to compare changes in approach. Also research can be done in a non intrusive manner, as work could be done on published material that is in the public domain. In addition content analysis can be used to analyze information created through other research techniques such as interviews and other material gathered for case studies, action research or narrative research.

Many research methods produce a large amount of material that can be reviewed and understood through the use of content analysis to identify major themes and underlying messages. Computerized systems assist to aid in content analysis and these are able to review large volumes of information and enable a number of ways of assessing information. They are limited in the ability to review the subtlety of language use however this can be overcome by the use of humans to code particular sections.

Content analysis is seen to provide information that can lead the researcher to draw inferences about the reasons why the information is portrayed in these ways. However, Kolbe and Burnet (1991) identify that it can also assist in theory development and they identified five roles for content analysis in theory development using Lijpharts (1971) categorizations for case study research. This research method can embellish, augment, accumulate and describe information. The need for systematic study and information acquisition, part of the initial steps in theory development, can be readily provided by content-analysis research.p.99.

Discourse analysis in business research

There appears to be significant diversity of views regarding the definition of discourse analysis. One view of discourse analysis, stemming from the philosopher H.P. Grice, sees it as a further development upon linguistic or semantic analysis,the study of the use of language as it flows or unfolds, as opposed to the rather atomistic sentence-based focus of stylistics or traditional linguistics (Sim, 1998, p. 231). As is, however, the case with much of the terminology utilised by postmodernist writers, the term "discourse", and consequently "discourse analysis" has a broader meaning in the context of the work of the seminal postmodern theorist Michel Foucault. "

A discourse, for Foucault, is the matrix of texts, the specialised languages and the networks of power relations operating in and defining a given field" (Sim, 1998, p. 242). More concisely, a discourse may be considered as "a well bounded area of social knowledge" (McHoul & Grace 1993, p. 31). The Foucaultian concept ofdiscourse challenges the basic premise that language conforms to, and indeed reflects, a rational, underlying objective pattern or structure of reality. Rather, the context, content and use of language, both spoken and written, is the consequence of political forces, and the dominant discourses which have emerged as a result of these political forces today define the reality within which we live.

Whilst discourse analysis may not constitute a stand alone research methodology, it provides a valuable tool by which the researcher can analyze the objectivity with which a given research question is approached, and may also provide a means for elucidating the manner by which social forces shape individual cognition, and subsequent collective actions. Within organizational research, particularly that which is concerned with the nature of organizational change, discourse analysis may prove to be a fruitful technique for uncovering the epistemology of the underlying assumptions which shape organizational culture and behaviour.

Discourse analysis can be used to trace the interconnection between a discourse and the social context in which it emerges. By analyzing a discourse critically alongside the social context, the method (of discourse analysis) can take account of agents and groups of agents who take part in discourse and trace the relations between discourse and socialprocesses (Garnsey and Rees, p. 1042)
Discourse analysis- selected readings

Hamilton , Sheryl N., Incomplete Determinism: A discourse analysis of cybernetic futurology in early cyberculture Journal of Communication Inquiry, Vol. 22, No. 2, April 1998, pp. 177-205

This article examines the terms cyberculture, and the ubiquitous present usage of cyber-hyphen-everything, from a Foucaultian perspective. It attempts to uncover the way that the use of cyber and associated terms and concepts influences our attitudes towards technology, knowledge, and the resultant power relationships between the cyber-savvy and those less so, leading to a particular view of the future which is gaining increasing currency.

Garnsey, Elizabeth, Rees, Bronwyn, Discourse and enactment: gender inequality in text and context, Human Relations, Vol. 49, No. 8, August 1996, pp. 1041-1065

This paper analyses the discourse associated with the UK Opportunity 2000 program -a government supported UK business initiative intended to increase the quality and quantity of womens participation in the workforce. The researchers examine how discourse influences individual cognition, in particular how the underlying assumptions and expectations which shape the discourse of equal opportunities make actually work against the espoused objectives of equal opportunity initiatives. The suggestion is made that "discourse analysis provides a tool for exposing the preconceptions and interests behind apparently neutral facades".

McHoul, Alec, Grace, Wendy, A Foucault Primer: discourse, power and the subject, Melbourne University Press, 1993

This work presents an overview of Foucault’s writings, particularly the relationships Foucault explored between discourse, power, and subjectivity. The placement of Foucaults concept of discourse (as opposed to the Non-Foucauldian conceptions of discourse is considered within the historical context of linguistic analysis and structuralism. In the opinion of this reviewer, a fairly demanding work, de'spite its claim to be an introductory account for the non-specialist reader.

Norris, Christopher, Deconstruction, theory and practice, Methuen, London, 1982

A comprehensive overview of the major players in the field of deconstruction, presented as a historical analysis of the development of deconstructionist philosophies as a response to structuralism. An in-depth discussion of the work of Jacques Derrida is presented. Of particular interest is a chapter detailing the critiques of deconstructive approaches, commencing with the ironic, and perhaps valid observation that deconstructionists do not, in general, tend to deconstruct their own works and interpretations.

Sim, Stuart (ed.) The Icon Critical Dictionary of Postmodern Thought, Allen & Unwin, Sydney, 1998

Any neophyte to the postmodernist world rapidly becomes overwhelmed by the vast array of new terminology, or worse still new meanings for "well known" terms and concepts which this field engenders. This work is therefore invaluable as a constant reference when pursuing other readings by post-modernist authors, who inevitably write for audiences assumed to have a deep and instinctive understanding of postmodernist jargon. The first section also provides a number of succinct overviews of the influence of postmodernism on modern culture and philosophy.

Content and discourse analysis - Discussion questions

1. Compare the traditional (quantitative/nonreactive) approaches to content analysis research with the newer (qualitative/reactive) approach ?
Business research must include a social or human relations side (that includes the industrial relation influences) and cultural analysis lends itself to this within organizations.

How could you apply this to your own organization in implementation and maintain the credibility also? How to get to a practical approach otherwise it would be dismissed as academic "gobbledegook"? The message, in order to get it accepted by the business world, might have to be adjusted Ð this means that it is not really to be presented in its pure research context but simply reviews outcomes - and the methodology has to be explained in terms of higher level components (for example, interviews are to be conducted but not that the outcome is to be analyzed using content or discourse analysis).

However, it will have to have a defined outcome that is identifiable in terms that those agreeing to the research in the organization will understand. This would also have to establish some sort of outcome for the research and conclusions that could be drawn, even though academically these would only be seen as suggestions rather than academic findings using the content analysis technique. Others who had done this type of research identified that the product delivered to the organization had to be translated into a report that made sense to the way they did business rather than providing the academic paper.

Note - It could be seen as esoteric (and therefore impractical or useless) as it is grounded in social research theory instead of fields associated commonly with organizational business research. Another way of using this approach could be to look at the well-trodden paths and the gaps between them and to look at the changes within the organizations to find some new insights. It was noted that there were functionalist paradigms existing in organizations. Research methods were not interested in following these lines, therefore the translation into the more accepted forms could be part of the research activity. Content and discourse analysis is to collect data and shape further work.

Discussion about using the instrument to achieve what purpose, and the necessity to discuss it with people in the organization.

2. Evaluate the ethical implications of using a combination of nonreactive and reactive approaches. One needs to consider the ethical issues when using reactive approaches in setting up and collecting the base information that will later be assessed using the nonreactive approach.

3. How can you overcome (a) the inherent bias and use of preconceptions in designing and analyzing the data ?

4. Can you validly infer outcomes from such analysis? Some researchers say you must use assessment of the total meaning to make sense of the answers.
It was noted that a reductionist approach would be rejected in systems theory. Taking information out of context of the complexity of the environment would reduce its meaning. Content analysis would be an issue, however discourse analysis would reveal the total picture. It needs to have an environmental context. Each might be appropriate for some purposes.

5. What are the dominant discourses in your chosen field of research. Why?

Both Content Analysis and Discourse Analysis can be used in a number of different types of research to analyze the data collected in the form of words or images, for example case studies, action research and narrative analysis. It was seen as a tool that in business would be useful as an adjunct to other forms of research.