Applied Quantitative Analysis in Education and the Social Sciences

To say that complex data analyses are ubiquitous in the education and social sciences might be an understatement. Funding agencies and peer-review journals alike require that researchers use the most appropriate models and methods for explaining phenomena. Univariate and multivariate data structures often require the application of more rigorous methods than basic correlational or analysis of variance models. Additionally, though a vast set of resources may exist on how to run analysis, difficulties may be encountered when explicit direction is not provided as to how one should run a model and interpret results. The mission of this book is to expose the reader to advanced quantitative methods as it pertains to individual level analysis, multilevel analysis, item-level analysis, and covariance structure analysis. Each chapter is self-contained and follows a common format so that readers can run the analysis and correctly interpret the output for reporting. Cover 1 Applied Quantitative Analysis in Education and the Social Sciences 4 Contents 8 Preface 10 Contributors 12 Part I Individual-Level Analysis 304 1 Extending Conditional Means Modeling: An Introduction to Quantile Regression 196 2 Using Dominance Analysis to Estimate Predictor Importance in Multiple Regression 47 3 I Am ROC Curves (and So Can You)! 78 Part II Multilevel Analysis 106 4 Multilevel Modeling: Practical Examples to Illustrate a Special Case of SEM 108 5 Linear and Quadratic Growth Models for Continuous and Dichotomous Outcomes 138 Part III Item-Level Analysis 182 6 Exploratory and Confirmatory Factor Analysis 184 7 Factor Analysis with Categorical Indicators: Item Response Theory 221 Part IV Covariance Structure Analysis 256 8 Introduction to Structural Equation Modeling 258 9 Latent Growth Curve Modeling Using Structural Equation Modeling 279 10 Latent Class/Profile Analysis 317 11 n-Level Structural Equation Modeling 342 Index 376 The mission of this book is to address advances in quantitative methods that are not covered in other texts, such as item response theory, the family of multilevel models (including cross-classified and growth), and mixture modeling.