Название: Configural Frequency Analysis: Foundations, Models, and Applications
Автор: Alexander von Eye, Wolfgang Wiedermann
Издательство: Springer
Год: 2022
Страниц: 401
Язык: английский
Формат: pdf (true), epub
Размер: 13.2 MB

This unique book provides a comprehensive and detailed coverage of configural frequency analysis (CFA), the most useful method of analysis of categorical data in person-oriented research. It presents the foundations, methods, and models of CFA and features numerous empirical data examples from a range of disciplines that can be reproduced by the readers. It also addresses computer applications, including relevant R packages and modules.

Configural frequency analysis is a statistical method that allows the processing of important and interesting questions in categorical data. The perspective of CFA differs from the usual perspective of relations among variables; its focus is on patterns of variable categories that stand out with respect to specific hypotheses, and as such, CFA allows for testing numerous substantive hypotheses.

The book describes the origins of CFA and their relation to chi-square analysis as well as the developments that are based on log-linear modeling. The models covered range from simple models of variable independence to complex models that are needed when causal hypotheses are tested. Empirical data examples are provided for each model. New models are introduced for person-oriented mediation analysis and locally optimized time series analysis, and new results concerning the characteristics of CFA methods are bolstered using Monte Carlo simulations.

In the following chapter, we introduce into the application of CFA in the context of the statistical programming environment R. In this environment, the packages cfa and confreq are available as well as an additional number of useful functions. The two packages can be downloaded free of charge from the Comprehensive R Archive Network (CRAN). Here, we focus on the application of the package confreq. Additional functions can be downloaded, also free of charge, from the web page. The package confreq can be installed from the Internet. In the following sections, we show how a number of CFA base models can be estimated in R. We begin with global CFA base models of zero, first, and second order. This is followed by application of Prediction CFA (P-CFA) and two-group CFA. This chapter concludes with an illustration of Bayes CFA.

Primarily intended for researchers and students in the social and behavioral sciences, the book will also appeal to anyone who deals with categorical data from a person-centered perspective.