Quantitative Applications in the Social Sciences – serie

Author William G. Jacoby focuses on graphical displays that researchers can employ as an integral part of the data analysis process. Such visual depictions are frequently more revealing than traditional, numerical summary statistics. Accessibly written, this book contains chapters on univariate and bivariate methods. The former covers histograms, smoothed histograms, univariate scatterplots, quantile plots, box plots, and dot plots. The latter covers scatterplot construction guidelines, jittering for overplotted points, marginal box plots, scatterplot slicing, the Loess procedure for nonparametric scatterplot smoothing, and banking to 45 degrees for enhanced visual perception. This book provides strategies for examining data more effectively. The resultant insights help researchers avoid the problem of forcing an inaccurate model onto uncooperative data and guide analysts to model specifications that provide accurate representations of empirical information.

This volume shows how odds ratios can be used as a framework for understanding log-linear models. Tightly organized, the book moves systematically from the paradigmatic 2x2 case to more complicated cases. Author Tamas Rudas carefully defines the odds ratio and demonstrates how it is a measure of association for tabular analysis. Throughout the book, Rudas provides real examples, including those from the famous Stouffer data on American soldiers in World War II. By presenting categorical data models in a unified framework, readers will learn not only how these models are related but also additional interpretations of these models not usually covered in other texts.

This volume provides readers with a simple, non-technical introduction to correspondence analysis (CA), a technique for summarily describing the relationships among categorical variables in large tables. It begins with the history and logic of CA, followed by an explanation of the analysis of large contingency tables and of survey data. The author shows readers the steps to the analysis: category profiles (relative frequencies) and masses (marginal proportions) are computed, the distances between these points calculated, and the best fitting space of n-dimensions located. In addition, the author provides glossaries on appropriate programs from SAS and SPSS for doing CA. The book concludes with a comparison of CA and loglinear models.

Researchers and professionals in sociology, market research, and other social sciences will find Applied Correspondence Analysis a useful tool for doing CA in the various statistical packages.

Filled with well-chosen empirical examples to teach the technique, this book provides a thorough guide to latent class scaling models for binary response variables. Substantive illustrations include a survey on academic cheating, children's mastery of spatial tasks, medical diagnosis of lung disease, attitudes toward the Army and behaviour during role conflict. The computer programs for latent class analysis are carefully reviewed and a website is offered for keeping abreast of the latest developments. This book will be of particular interest to those doing detailed scale analysis in psychology, sociology and education.

Neural Networks have influenced many areas of research but have only just started to be utilized in social science research. Neural Networks provides the first accessible introduction to this analysis as a powerful method for social scientists. It provides numerous studies and examples that illustrate the advantages of neural network analysis over other quantitative and modeling methods in wide spread use among social scientists. The author presents the methods in an accessible style for the reader who does not have a background in computer science. Features include an introduction to the vocabulary and framework of neural networks, a concise history of neural network methods, a substantial review of the literature, detailed neural network applications in the social sciences, coverage of the most common alternative neural network models, methodological considerations in applying neural networks, examples using the two leading software packages for neural network analysis, and numerous illustrations and diagrams.

This introductory guide to using neural networks in the social sciences will enable students, researchers, and professionals to utilize these important new methods in their research and analysis.

This handy guide gives the novice researcher a clear description of the standard tools of the trade. Unlike some texts which focus on either design or statistics, this book covers the fundamentals of design, together with experiments and observational methods. There is an exposition of major tests of significance with formulas plus easy verbal interpretations, and "boxes" embedded in the text contain prototypic applications.

Researchers have been complaining about the lack of one single place to find information on computing effect sizes in analysis of variance (ANOVA), until now. Authors Jose M. Cortina and Hossein Nouri begin with a literature review of previous treatments of the topic (including corrections to the misleading treatments of repeated measures and ANCOVA (analysis of covariance) designs). They introduce the effect sizes, by defining the term and explaining how it is computed from summary and test statistics for the simple two independent group design. They next provide a description of methods for computing effect sizes from the results of one-way designs with more than two groups, and then extend these methods to cases in which the effects of interest are embedded within the context of two and three-way ANOVA's. They conclude the book with an explanation of the methods for computing effect size from the results of ANCOVA designs followed by the methods for computing effect sizes from the results of repeated measures. Throughout the book, the authors offer examples with worked-out computations to illustrate each technique. Researchers who need to estimate their effect size of run a meta-analysis will find this book very useful.

While regression analysis traces the dependence of the distribution of a response variable to see if it bears a particular (linear) relationship to one or more of the predictors, nonparametric regression analysis makes minimal assumptions about the form of relationship between the average response and the predictors. This makes nonparametric regression a more useful technique for analyzing data in which there are several predictors that may combine additively to influence the response. (An example could be something like birth order/gender/and temperament on achievement motivation).

Unfortunately, researchers have not had accessible information on nonparametric regression analysis, until now. Beginning with presentation of nonparametric regression based on dividing the data into bins and averaging the response values in each bin, Fox introduces readers to the techniques of kernel estimation, additive nonparametric regression, and the ways nonparametric regression can be employed to select transformations of the data preceding a linear least-squares fit. The book concludes with ways nonparametric regression can be generalized to logit, probit, and Poisson regression.

- Provides an accessible but thorough introduction to GLM, exponential family distribution, and maximum likelihood estimation- Includes discussion on checking model adequacy and description on how to use SAS to fit GLM- Describes the connection between survival analysis and GLM

 This book is an ideal text for social science researchers who do not have a strong statistical background, but would like to learn more advanced techniques having taken an introductory course covering regression analysis.

This book builds on John Fox's previous volume in the QASS Series, Non Parametric Simple Regression. In this monograph readers learn to estimate and plot smooth functions when there are multiple independent variables. While regression analysis traces the dependence of the distribution of a response variable to see if it bears a particular (linear) relationship to one or more of the predictors, nonparametric regression analysis makes minimal assumptions about the form of relationship between the average response and the predictors. This makes nonparametric regression a more useful technique for analyzing data in which there are several predictors that may combine additively to influence the response. (An example could be something like birth order/gender/and temperament on achievement motivation).

Unfortunately, researchers have not had accessible information on nonparametric regression analysis, until now. Beginning with presentation of nonparametric regression based on dividing the data into bins and averaging the response values in each bin, Fox introduces readers to the techniques of kernel estimation, additive nonparametric regression, and the ways nonparametric regression can be employed to select transformations of the data preceding a linear least-squares fit. The book concludes with ways nonparametric regression can be generalized to logit, probit, and Poisson regression.

The focus in this Second Edition is again on logistic regression models for individual level data, but aggregate or grouped data are also considered. The book includes detailed discussions of goodness of fit, indices of predictive efficiency, and standardized logistic regression coefficients, and examples using SAS and SPSS are included.

Updated coverage of unordered and ordered polytomous logistic regression models.