In lieu of an abstract, here is a brief excerpt of the content:

Cooper, H., and Hedges, L. V. (Eds.) 1994. The Handbook ofResearch Synthesis. New York: Russell Sage Foundation 29 META-ANALYTIC SYNTHESIS FOR THEORY DEVELOPMENT NORMAN MILLER University of Southern California VICKI E. POLLOCK University of Southern California CONTENTS 1. Introduction 2. Defining Theory 3. Defining Theory Testing 4. Meta-Analysis for Theory Testing 4.1 Cause and Effect 4.1.1 When meta-analysis can provide causal evidence 4.1.2 Why most meta-analytic results do not provide causal evidence 4.2 Explanatory Tasks Within Meta-Analysis 4.2.1 Assessing a relationship and ruling out artifactuality 4.2.2 Assessing the form of a relationship 4.2.3 Assessing the critical components of theoretical concepts 4.2.4 Assessing the scope or limits of a relationship 4.2.5 Specifying mediating processes 4.2.6 Accounting for the variance among study outcomes 4.3 Advantages of Using Meta-Analysis to Test Theory 4.3.1 The power of quantitative integration 4.3.2 Avoiding the logistical constraints of the single study 4.3.3 Expanding temporal horizons 4.3.4 Avoiding the constraints imposed by changes in ethical and institutional standards 5. Three Types of Meta-Analytic Evidence in Tests of Theory 5.1 Type A Tests: Aggregation of Study-Level Evidence 5.2 Type B Tests: Differentiation and Generalization 5.3 Type C Tests: Evidence Bearing on New Theories 458 458 459 460 460 461 461 462 463 463 463 465 466 466 466 467 467 467 467 468 468 468 468 457 458 TYING RESEARCH SYNTHESIS TO SUBSTANTIVE ISSUES 5.4 Type A, B, and C Tests in Published Meta-Analyses 6. Other Issues in Testing Theory via Meta-Analysis 469 469 476 478 6.1 Procedures for Comparative Meta-Analytic Theory Testing 6.2 Meta-Analysis and the Psychology of Science 7. References 1. INTRODUCTION In this chapter we contend that some of the features of meta-analysis make it superior to individual studies for testing and formulating theories. In section 2 we discuss the ambiguous use of the term "theory." In section 3 we discuss what we mean by "testing theory." In section 4 we review the potential roles of meta-analysis in developing and evaluating theory, which include assessing causal relationships as well as fulfilling other explanatory tasks. In this section we also consider its advantages relative to single empirical studies. In section 5 we discuss three types of meta-analysis that differ in the way and degree to which they promote theoretical advance. Finally, we discuss issues that confront the meta-analyst who undertakes comparative theory testing . 2. DEFINING THEORY A hypothesis is a guess at an empirical relationship that links two theoretical concepts. An empirical law is a confirmed hypothesis-a relationship between two concepts that has been (relatively) well established by research . A scientific theory is a framework for interpreting a conceptually related set of empirical laws. As an abstract nomological network, a theory also provides a basis for deriving new hypotheses from its more general postulates, which can then be tested by research. In Lewin's metaphor (1951): "The ultimate goal is to establish a network of highways and superhighways, so that any important point may be linked with any other. This network of highways will have to be adapted to the natural topography of the country and will thus itself be a mirror of its structure and of the position of its resources" (p. 3; compare Hempel 1952 for a similar description). In addition, a theory must have scope. In principle, the collection of laws describing the behavior of a lever as a function of various magnitudes of weights, placed at different distances from the fulcrum on which the lever is positioned, can be termed a theory. Innumerable 479 individual laws, including the law of equilibrium, each based on a specific combination of weights, and distances , will predict the direction and speed of the lever's rotation and can be derived from more general principles . Most scientists, however, would balk at terming all this "the general theory of the lever," objecting that it lacks sufficient scope. Instead, they might prefer to view the effects of the specific combinations of weights and distances as a single general law. To scientifically explain something is to state the laws known about it. When one theory is said to provide more or less explanation or understanding than another, what is being referred to is the relative number and scope of the laws...

Share