190 likes | 243 Views
Alan Bryman. Social Research Methods. Chapter 13: Content analysis. Slides authored by Tom Owens. Approach to analysis of documents and texts Quantifies content in terms of predetermined categories Systematic and reliable Usually a quantitative research strategy
E N D
Alan Bryman Social Research Methods Chapter 13: Content analysis Slides authored by Tom Owens
Approach to analysis of documents and texts Quantifies content in terms of predetermined categories Systematic and reliable Usually a quantitative research strategy Different from semiotics and ethnographic content analysis What is content analysis? Pages 289-291
Objective, systematic and quantitative description of the manifest content of communication (Berelson, 1952) Focus on mass media communications Objective, systematic identification of specified characteristics of messages (Holsti, 1969) Includes latent content Definitions of content analysis Page 289
Must be clearly specified before analysis Decide which dimensions of texts to quantify e.g. news coverage of an issue: who, what, how much, where and why? How many different views represented? What is selected for inclusion? Omissions: what is not reported? Does the amount of coverage change over time? Research questions Pages 291
Selecting a sample • If working with the mass media (most likely case), two dimensions are possible for random sampling: • The media itself • Dates of publication
Progressive narrowing down Which types of text? Printed or visual data? Documents? Mass media? If mass media, which kind? TV, radio, newspapers, magazines… more than one type? For each type of text, which examples? e.g. tabloid or broadsheet newspapers Sampling media Page 293
Starting dates may be predetermined by an historical event (like the terrorist attack on the World Trade Centre on 11 September 2001) More open if the study is an ongoing, general phenomenon End dates can be a matter of judgement So, stick to the principles of probability sampling (as outlined in Chapter 8) Sampling dates Pages 293, 294
Significant actors: protagonists and alternative voices Words: frequency of words or phrases (e.g. ‘hooligan’ or ‘non-smoker’), using computer software for analysis (e.g. Wordsmith) Subjects and themes Dispositions: values, bias and ideology What is to be counted? Pages 295-298
A tabular form onto which coded data will be entered Each column represents a dimension to be analysed Each row represents a unit of analysis (item of text) Codes are written into blank cells in table The coding schedule Page 298
Example of a coding schedule Figure 13.1 Page 291
A set of instructions for coders Lists all possible categories for each dimension Shows which codes/numbers refer to which category Gives guidance on how to decide on a code Explains what to do if more than one code applies May use an existing content analysis dictionary as a starting point. The coding manual Page 299
Portions of a coding manual From Figure 13.2 Page 300
Ensure that coding scheme has: discrete dimensions mutually exclusive categories exhaustive categories clear instructions to coders a clearly specified unit of analysis (distinction between the media and the event reported) Pilot the study to make sure of consistency between coders (inter-coder reliability) and consitency over time for each coder (intra-coder reliability) Avoiding potential pitfalls Pages 303, 304
Transparency Easeof longitudinal analysis Unobtrusiveness Flexibility Ease of access Advantages of content analysis Pages 304, 305
Questions of authenticity, credibility and representativeness of documents Interpretation by coders Invalid inference of manifest content Inability to answer `why?’ questions Has an atheoretical approach Disadvantages of content analysis Pages 306, 307