Search

A single-case design is an experimental design for a study in which one entity is observed repeatedly during a certain period under different levels of at least one independent variable. Single-case designs have a long tradition and growing impact in behavioral science. They are the designs of first choice if demonstration of an effect in a single case is aimed at or if a test of a theory in a single case is sufficient (or the only possibility), and they are most compatible with clinical practice. The systematic and purposive planning of interventions in single-case research represents the main difference with both observational time series research and qualitative case study research. In addition, single-case research can gain internal and statistical-conclusion validity by including randomization in the design. Two types of randomized single-case designs are presented: randomized alternation designs and randomized phase designs. Finally, external validity of single-case results can be demonstrated or tested by simultaneous or sequential replication and the corresponding simultaneous randomization tests and meta-analytic procedures. Keywords: experimental design; time series; case study; random assignment; alternating treatments design; multiple baseline design; reversal and withdrawal design

A meta-analysis combines or compares in a systematic and quantitative way the results from several studies in which a similar research question is investigated. If differences in research outcomes are entirely due to sampling variation, combining the outcomes results in more precise answers to the research questions. In case of systematic inconsistencies in study outcomes, a meta-analysis offers the possibility to look for an explanation for the heterogeneity, by investigating whether study outcomes are related to one or more study characteristics. In this article, the historical and theoretical context of meta-analysis and the successive steps in a meta-analysis are described shortly, and several models, techniques and caveats are discussed. Keywords: meta-analysis; heterogeneity of study outcomes; mixed effects models; random effects models; file drawer problem