Content Analysis and Thematic analysis

• Define content analysis

  -type of observational research in which people are studied indirectly via the qualitative communications they have produced.
• What is coding and quantitive data

  -This is the first stage of content analysis.

  -Some pieces of data are extremely large so there is a need to categorise the information into meaningful units.
• What are some examples of coding

  -counting up the number of times a particular word appears in a text or looking at the way in which gender roles are shown in adverts.
• Define frequency analysis

  -the number of times certain words come up (e.g. number of times that sexual references come up in day-time television)
• Define concordance analysis

  -the number of times certain phrases come up.
• Define thematic analysis

  -This is the process of coding and identifying themes which are closely linked
• Themes are more likely to be

  -descriptive than coding units
• Once the researcher is satisfied that the themes they have developed cover most aspects of the data

  -they may collect a new set of data to test the validity of the themes/categories.
• Assuming that these explain the new data adequately,

  -the researcher will write up a final report, often using quotes from the data to illustrate each theme.
• What is the first step in conducting content analysis

  -data is collected
• step 2

  -researcher reads through or examines data making themselves familiar with it
• Step 3

  -researcher identifies coding units
• step 4

  -data is analysed by applying the coding units
• step 5?

  -a tally is made up of the number of times a coding unit appears
• What are some strengths of content analysis

  -Content analysis is highly flexible in that it produces both quantitative & qualitative data.​

  -Great to study emotions and motivation​

  -Can establish what caused the behaviour
• What is another strength of content analysis

  -Content analysis can ‘get around’ most ethical issues usually associated in research as much of the material that they may use is already within the public domain.

  -They may however need to obtain permission to use the data.
• What are some weakness of content analysis

  -Not Scientific (I.e. it’s hard to statistically prove anything)​

  -Can’t really generalise​

  -Reliability - how one person views the report/data may be different from someone else​

  -Validity - the language is ambiguous and may be misinterpreted.
• What is another weakness of content anaylsis

  -As people are studied indirectly the data is analysed out of context

  -lack of objectivity
• What are 2 ways the reliability of content analysis can be improved

  -Researchers should clarify/establish/agree/review operationalised codes/categories etc​

  -Researchers could create new categories if necessary (that are mutually exclusive)
• 2 more ways

  -Researchers should be trained carefully in using these identified codes/categories (words/phrases)​

  -They should analyse a small number of diary extracts using the same analytic/coding system
• any improvement in reliability

  -could be determined by establishing inter-rater/inter-observer reliability (between the two researchers); test-retest reliability.
• The two researchers results are

  -compared for similarity to check improvements. Statistical analysis is performed and if the correlation coefficient is 0.8 or above, then there is high reliability.
• Outline how test retest reliability is carried out in the context of content analysis

  -content analysis repeated on a second occasion using the same interview data ​

  -compare the results of the two separate analysis (number of occurrences of each) ​

  -researchers could calculate the correlation between the two ratings ​

  -researchers generally accept 0.8 correlation between the test and the re-test.
• Outline how inter-rater reliability is carried out in the context of content analysis

  -use a second person to work with the original researcher ​

  -they could read the interviews (separately) and devise a set of categories (and agree operational definitions) ​

  -they could tally the occurrences of each of the categories of the interviews (separately) ​

  -they could compare their tally charts looking for agreement

  -calc correlation and if it is 0.8 accept