Analysis of Qualitative Data

Analysis of Qualitative Data: Volume 1, Introductory Topics introduces log-linear models oriented toward the social scientist, including assessments of the variability of parameter estimates using algebraic equations and summation notation. The book also contains examples involving basic problems in survey research, such as memory error. Other examples pertain to the General Social Survey of the National Opinion Research Center that examines public opinion on abortion, as well as the variations in homicide rates related to variables (such as race or sex of victim). The text explains the quantitative assessment of the size of departures from independence of polytomous variables by investigating the linear combinations of log cells means, known as log cross-product ratios. The book discusses the use of log odds, conditional log odds, cross-product ratios, and conditional cross-product ratios in interpreting hierarchical models such as those found in the General Social Surveys. The text describes logit models, namely, the Newton-Raphson algorithm used to explore the relationship of a dichotomous dependent variable to one or more independent variables. The book can serve and benefit mathematicians, students, or professor of calculus, statistics, and advanced mathematics.

Preface

Acknowledgments

Contents of Volume 2

1 Polytomous Responses

1.1 Effects of Memory—An Introduction to Log-Linear Models

1.2 Self-Classification by Social Class

1.3 Suicide and the Poisson Distribution

1.4 Disbibution of Reported Suicides by Region

1.5 General Techniques for Log-Linear Models

Exercises

2 Complete Two-Way Tables

2.1 A Two-by-Two Example

2.2 A 2 x s Table: Ethnicity and Women’s Role Attitudes

2.3 An r x s Table: Psychiatric Therapy and the Psychoses

2.4 Comparison of Samples

2.5 Additive Log-Linear Models for Poisson Observations

2.6 Additive Log-Linear Models

Exercises

3 Complete Three-Way Tables

3.1 Race, Sex, and Homicide Weapon—A 2 x 2 x 2 Table

3.2 Woman’s Place, Sex, and Education

3.3 Possible Hierarchical Models for Three-Way Tables

3.4 Computational Procedures

3.5 Selection of Models

Exercises

4 Complete Higher-Way Tables

4.1 Comparison of Three-Way Tables: Year of Survey, Sex, Education, and Place

4.2 Year of Survey, Religion, Education, and Attitudes Toward Nontherapeutic Abortions

Exercises

5 Logit Models

5.1 Logits and Table 3.1

5.2 Logits and Women's Roles

5.3 Logits, Women's Roles, and Small Frequency Counts

5.4 Properties of Logit Models

5.5 Quantal-Response Models

Exercises

References

Example Index

Subject Index