Harnessing Available Evidence in Single-Case Experimental Studies: The Use of Multilevel Meta-Analysis.

Area of Science:

• Behavioral Science
• Research Methodology
• Statistics

Background:

• Multilevel models have been used for meta-analyses of single-case experimental design (SCED) studies for approximately 20 years.
• The application of these models has grown, with increasing complexity and empirical evaluation.

Purpose of the Study:

• To provide a state-of-the-art overview of multilevel models for SCED research.
• To summarize simulation findings and discuss ongoing challenges in the field.

Main Methods:

• Review of basic and advanced multilevel models for SCED data.
• Summary of empirical evaluations of multilevel approaches.
• Discussion of methods to address SCED complexities.

Main Results:

• Multilevel models can effectively integrate and compare results from multiple SCED studies.
• These models can accommodate various complexities inherent in SCED data, including autocorrelation, time trends, and heterogeneity.
• Extensive research has validated the utility of multilevel models in diverse SCED research scenarios.

Conclusions:

• Multilevel modeling is a sophisticated and increasingly utilized approach for synthesizing SCED research.
• The flexibility of these models allows for the analysis of complex SCED data structures.
• Further research is needed to address remaining issues in the application of multilevel models to SCED.