Content Analysis
Also known as: text coding, message analysis, 内容分析.
A method that turns texts or images into countable categories using explicit coding rules, so that claims about content can be checked and repeated.
Synthetic dataExamples on this page use synthetic data created for illustration.
Not adviceThis is educational material. It is not investment, business or other professional advice.
Definition
Content analysis is a research method that turns text, images or other messages into countable categories using explicit coding rules, so that claims about the content can be checked and repeated. Its reliability is shown by how well independent coders agree.
In plain terms
The steps are: decide what the unit is (an article, a paragraph, a sentence), define the categories in a written codebook, train coders, have two or more of them code the same sample, measure their agreement (see Cohen's Kappa), and only then code the rest and count.
Automatic methods, such as word lists and classifiers, are used for scale, but they must be validated against human coding of a sample from the same material. Counting words is not reading meaning.
Example
Two coders label the tone toward “Acme” in 10 articles (positive, neutral, negative):
| Outcome | Articles |
|---|---|
| Both say positive | 3 |
| Both say neutral | 3 |
| Both say negative | 2 |
| They disagree | 2 |
Raw agreement is 8 of 10 = 80%. The two disagreements are the useful part: they show which codebook rule is ambiguous and needs rewriting before the remaining articles are coded.
Common misunderstandings
- “Content analysis is counting words.” It is counting coded categories defined by rules that capture meaning.
- “Automatic coding removes the need for human coders.” It needs human-labelled samples to check that it works.
- “High agreement shows the coding is correct.” Coders can agree and still be wrong together; agreement shows the rules are clear.
References
- Klaus Krippendorff. Content Analysis: An Introduction to Its Methodology . SAGE Publications, 2018. ISBN 978-1412983150.
- Justin Grimmer, Brandon M. Stewart. Text as Data: The Promise and Pitfalls of Automatic Content Analysis Methods for Political Texts . Political Analysis, 2013. DOI: 10.1093/pan/mps028