Content analysis is a research tool used to determine the presence of certain words, themes, or concepts within some given qualitative data (i.e. text).
What is Content Analysis?
‘An approach to the analysis of documents and texts (printed or visual) that seeks to quantify content in terms of predetermined categories and in a systematic and replicable manner’ (Bryman, 2012)
It is an objective and systematic approach suited to certain types of research question and which is amenable to quantification.
Related: Read more on Qualitative data analysis
What type of research is content analysis suited to?
Frequency counts of manifest topics dealt with in different media. Could be frequency of:
- Coverage of specific issues (e.g. Brexit,)
- Specific terminology (e.g. ‘Luxury’)
- Prominence (e.g. What page of a newspaper a specific issue is dealt with or how often – e.g)
Frequency counts of latent themes emerging; this involves some level of interpretation of meaning (Thematic analysis) For instance:
- media attitudes to issues (e.g. Brexit)
- values implied in use of specific terminology (e.g ‘Luxury’)
So content analysis can overlap with discourse or linguistic analysis, semiotics (study of signs and symbols)
Content analysis can also be applied to research beyond the linguistic… e.g
- Photo images in media
- Song lyrics
- Logos, signs and branding
- Advertisements
- Speeches
- Blogs
- Official documentation (e.g. meeting minutes, annual reports, policies / strategies)
- Social media…
Sampling
- Will be ‘purposive’ sampling – i.e. dependent on the researchers perspective on what is relevant to the purpose of the research (e.g what type of newspapers, web sites, official documents etc)
- Dates will be of importance and again specific to the research project. The extent, frequency of when the sample is drawn will be largely dependent on the occurrence of the phenomena.
- Probability sampling is usually employed when the population (e.g. range of newspapers) is large and the date range is extensive, but can proceed along the lines of selecting a start date and then every nth day
What is counted?
It depends on the research project:- Method refers to common units of analysis and these are –
- Significant actors (who produced, who is focus, who acts as alternative?)
- Words – the frequency of specific relevant words used (e.g. ‘Luxury’, ‘brexit’) Can use software if documents are available as e-documents
- Subjects and Themes – coding text in terms of emergence of subjects and themes of specific relevance to the research project (n-vivo useful here). Manifest and latent issues here.
- Dispositions (Sentiment) – e.g. Positive vs Negative toward a specific view on the subject…again involves interpretation and manifest / latent issues. Can involve coding, especially around specific ideologies, beliefs or discourses
Coding
- Involves the generation of specific codes (alpha-numeric – e.g. A1,2,3,B1,2,3 etc) to represent different subjects, themes, phenomena
- Coding schedule is a specific form designed for the research onto which the data relating to the codes are entered. Coding manuals might be involved for larger projects giving guidance on using the codes in turning specific instances into data
Advantages of Content Analysis
- Transparency – it is easy to set up and is replicable
- Longitudinal studies possible – can access clearly dated past documentation to establish frequencies
- It is unobtrusive – no human research subjects involved (Privacy and consent a consideration though)
- It is flexible and can be applied in different settings
- Can get information on groups that cannot be accessed for participation
Disadvantages of Content Analysis
- Can only be as good as the documentation or whatever material being analysed – issues of authenticity, credibility and representativeness
- Coding consistency can be a problem
- Assigning latent meaning of content
- Not good at ‘why?’ questions – causation
- Atheoretical – focus more on what is measurable rather than significant
Related: Complete research process explained
Leave a Reply