Content analysis

What is content analysis?

A research method which involves counting communication phenomena and categorization them according to a taxonomy or typology scheme.

Especially useful for archived data (texts, transcripts, videotapes, etc.)

Allows for unobtrusive measures in many cases

"Safe" form of data analysis--if you mess up you can start all over again

What can be content analyzed?

Any episode of communication that can be videotaped or transcribed, broken down into discrete units, and codified or categorized can be subjected to content analysis

real or simulated conversations

newspaper stories

language styles

compliance gaining strategies

group interaction

Steps in content analysis

1. Identify the data set to be used

typically artifacts, videotapes, transcripts, tape recordings, etc.)

2. Develop or borrow a taxonomy (the categories into which the data will be coded)

categories must be valid: isomorphic with the real-life communication

mutually exclusive: separate, distinct categories with no overlap

exhaustive: all the data fits into some category

3. Determine the unit of analysis, called "unitizing"

example: what constitutes an argument

example: marital comm: one person or two?

4. Train coders in the use of the coding scheme

5. Code the messages into the categories

inter-rater reliability matters: typically 80-90%

6. Analyze the data and interpret the results

Additional Considerations

process can be time-consuming

necessity of transcribing, coding data

better suited for examining "manifest" as opposed to "latent" meaning

loses some of the richness inherent in the message

inductive versus deductive typologies

                    advantages and disadvantages of each