Information theory : structural models for qualitative data /
Krippendorff introduces social scientists to information theory and explains its application for structural modeling. He discusses key topics such as: how to confirm an information theory model; its use in exploratory research; and how it compares with other approaches such as network analysis.
| Основен автор: | Krippendorff, Klaus. |
|---|---|
| Формат: | Електронна книга |
| Език: | English |
| Публикувано: |
Beverly Hills, Calif. :
Sage Publications,
℗♭1986.
|
| Серия: |
Quantitative applications in the social sciences ;
no. 07-062. |
| Предмети: | |
| Онлайн достъп: |
http://search.ebscohost.com/login.aspx?direct=true&scope=site&db=nlebk&AN=24722 |
| Подобни документи: |
Print version::
Information theory. |
Съдържание:
- Foreword
- 1. Qualitative data
- 2. Selective information
- 3. Entropy, diversity, variety
- 4. Shannon's theory of communication
- Noise and equivocation
- Information transmitted
- Redundancy
- 5. Comparisons of qualitative variates
- Informational distance
- Informational bias
- 6. Structural models
- Parameters
- Composition
- Interactions
- Relations between models: descendency
- Lattices
- 7. Models with and without loops
- 8. Information in models and in data
- 9. Structural zeros
- 10. Degrees of freedom
- 11. The significance of information quantities
- 12. Maximum entropy computations
- Models without loops and without structural zeros
- Models with loops or with structural zeros
- 13. Confirmation
- The goodness of fit of a model
- The amount of information modeled
- The complexity of a model's components
- The contributions a component makes
- The strength of relations (association)
- The amount of interaction
- Strata within models
- 14. Exploration
- Searching for the ordinalities of appropriate models
- Searching for optimum models
- Algebraic techniques
- 15. Comparisons with alternative approaches
- Network and path analyses
- Chi-square
- Analysis of variance
- Log-linear modeling
- The most basic reference possible.