THE HANDBOOK OF RESEARCH METHODS IN SOCIAL AND PERSONALITY PSYCHOLOGY: A Toolbox for Serious Researchers

Social and personality psychologists have always embraced a creative diversity of methods, ranging from participant observation (Festinger, Riecken, & Schachter, 1956) to psychometric test construction (Osgood, Suci, & Tannenbaum, 1957), using verbal, behavioral, and physiological measures in laboratory and field settings. We have invented new research designs (Campbell & Stanley, 1966) and new statistical analyses (Campbell & Fiske, 1959; Kenny & La Voie, 1984). We have discovered sources of error that are common to all research on humans (Orne, 1962; Rosenthal, 1963), and developed techniques to overcome them. Personality psychologists are most famous for creative contributions to psychometric theory that solve measurement problems inherent in self-report and observer data. Social psychologists are most famous for their ingenious laboratory experiments, in which significant social psychological phenomena are re-created in simplified form in a setting that eliminates confounds and extraneous noise. When familiar methods seemed wrong for the research questions, social and personality psychologists happily devised new ones; methodological elegance and wit were admired nearly as much as theoretical insight. With new questions and new technology, the variety of methodological choices continues to proliferate, and the task of compiling a comprehensive handbook of research methods in social and personality psychology is a daunting one. Reis and Judd’s aim is to “demonstrate . . . the tremendous methodological richness and innovativeness to be found in social psychological research” and “to provide social-personality psychologists with resources for expanding the methodological diversity employed in their research” (pp. xi–xii). On the whole, the book achieves these goals. The range of methods discussed is considerably broader than the coverage of most textbooks on research methods in social and personality psychology; the handbook includes chapters on computer simulation (Hastie and Stasser), physiological measures (Blascovich), surveys (Visser, Krosnick, and Lavrikas), content and narrative analysis (Charles Smith), behavioral and observational coding (Bakeman), techniques for assessing growth and change (Collins and Sayer), and meta-analysis (Johnson and Eagly). The metaphor of the toolbox recurs again and again, along with the praiseworthy exhortation to use multiple methods. The book is divided into three sections: Design and Inference Considerations, Procedural Possibilities, and Data Analytic Strategies. The design chapters address the usual issues of internal and external validity; replication; experiments, quasi-experiments, and nonexperiments; and holding variables constant versus varying them systematically versus letting them vary randomly. These chapters are clear, intelligent expositions and extensions of the work of Campbell and others, and should be very useful for a student who has decided on a basic procedure and wants to use it most effectively. (Hastie and Stasser’s chapter on computer simulation, included in this section, is quite different: It is a step-by-step exposition of how to use the method to test theory.) The section on procedural possibilities is almost exclusively devoted to various kinds of measures, and the section on data analysis focuses on various kinds of statistical procedures. In the sections on procedures and analyses, most of the authors describe a particular kind of measurement or analysis—a particular tool—urge that it be added to the toolbox, and give a more or less detailed description of how to use it. The editors clearly took some pains to achieve consistency of coverage across the chapters. Most of the chapters describe the relevant threats to internal and external validity, most of the chapters on design and measurement mention data-analytical considerations as well, and most authors make the arguments (albeit fairly abstractly) that (a) what is most important is to choose a method that fits the question and (b) it is better to design and conduct the study right in the first place than to rely on elaborate statistics to patch things up later on. Most of the chapters are admirably broad in their outlook, discussing a range of methods within the assigned domain. The chapters by Bakeman on behavioral observation and by Kerr, Aronoff, and Messé on methods of small-group research are particularly impressive in this regard. Other chapters are not as broad in their coverage. The one by Bargh and Chartrand, billed in the preface as a chapter on cognitive mediation, is narrowly focused on priming and automaticity; Bartholomew, Henderson, and Marcia’s chapter on interviewing is heavily focused on a particular type of coded semistructured interview that, as they acknowledge, is labor intensive and appropriate for a particular kind of research question. The book as a whole, despite its title, devotes far more attention to the concerns of social psychologists than to those of personality psychologists. Individual differences are mentioned in several chapters, but are not the major focus of attention, and personality-test construction shows up only in John and Benet-Martínez’s chapter on measurement.