An Introduction to Content Analysis

1
An Introduction to Content Analysis

• Communication Research
• Week 9Myra Gurney

2
Definition

• Aims to identify specific characteristics of
  communications systematically and objectively in
  order to convert the raw material into scientific
  data (Mostyn, 85)

3
What is content analysis?

• Text-based quantitative analysis
• Systematic and consistent analysis of textual
  expression
• Transforming text into numbers
• Understanding textual patterns quantitatively
• Generating a conceptual understanding of textual
  elements
• Tracing ideas

4
What is content analysis?

• Underlying concepts
• Text as data
• Words as indicators of ideas
• Patterns in text form basis for inferences
• Basic assumptions
• Coders are representative of an audience
• Textual meaning is consistent across documents
• Patterns in words represent patterns in concepts

5
Human versus computer coding

• Human coding
• Useful for complex coding concepts
• Personal judgment a part of coding process
• Takes time and money
• Computer coding
• Automated text search
• Useful for large datasets
• Context irrelevant
• Personal judgment not part of coding process
• Requires more extensive code testing
• Cheaper and faster

6
Data

• Textual examples
• Newspaper articles
• Headlines
• Paragraphs
• Length, words dedicated to a subject, tone, bias,
  etc.
• Political candidate speeches eg PM Howard
• Mentions of ideas, tone, introductions, closings,
  etc.

7
Image examples

• Television ads
• Published images of politicians

8
Sampling

• Sample of dataset
• E.g., all articles on Lewinski scandal
• Sample of possible data
• E.g., articles in a major newspaper New York
  Times articles between 1990 and 2000 becomes
  dataset

9
Analytical Concepts

• Codes
• Codes should capture ideas you want to analyze
• Developing codes that correspond to theory
• E.g., references to religion in presidential
  candidate speeches God, religion, Prayer,
  faith, religious ideas

10
Analytical Concepts

• Unit of analysis
• Document
• E.g., Newspaper article, speech, advertising spot
• Paragraph
• Consistent identification of paragraphs
• Questionable in speech transcripts
• Useful for newspapers or other published document
• Conceptual units
• Units that correspond to theory
• E.g., Introduction, argument, explanation

11
Developing Codes

• Theory-based
• Codes must correspond to theoretical purpose of
  project
• Identify patterns
• Discursive patterns
• E.g., use of the word God, appearance of
  references to religion
• Grammatical patterns
• E.g., passive voice, imperative statements
• Conceptual patterns
• E.g., Strategic versus issue focused language
• Exhaustive categories

12
Analytical Concepts

• Coding sheets
• Develop the instrument
• Simplicity is the key!

13
Analytical Concepts Inferences

• Inferences permit conclusions
• Validity
• Most often validity only extends to dataset
• Random samples of sample permit conclusions about
  whole dataset
• Difficult to infer patterns among other datasets
• E.g., coding ads from Procter Gamble does not
  permit inferences about all ads
• Assumptions about inferences
• Assuming that your codes are related to your
  question!

14
Analytical Concepts Error analysis

• Coding problems
• Coders might be unique
• Codes might not correlate with theory/ inferences