Location-Scale Meta-Analysis and Meta-Regression as a Tool to Capture Large-Scale Changes in Biological and...

Area of Science:

• Ecology
• Evolutionary Biology
• Meta-analysis

Background:

• Ecological and evolutionary meta-analyses often assume constant heterogeneity (homoscedasticity) across studies.
• This assumption can obscure important patterns related to the variability of effect sizes.
• Understanding changing heterogeneity is crucial for interpreting biological responses to global environmental change.

Purpose of the Study:

• To introduce and develop a location-scale meta-analysis and meta-regression framework.
• To model both the mean (location) and variance (scale) of effect sizes, explicitly accommodating heteroscedasticity.
• To reveal when and why heterogeneity changes, enhancing biological interpretability.

Main Methods:

• Developed a location-scale meta-analysis and meta-regression framework.
• Incorporated hierarchical (multilevel) structures and phylogenetic relationships.
• Extended concepts of relative heterogeneity and publication bias to the scale part of meta-regression.

Main Results:

• The framework models both the location and scale of effect sizes, revealing patterns previously obscured by homoscedasticity assumptions.
• Demonstrates how study design, measurement, environment, and evolutionary history can systematically shift variance.
• Identifies when and why heterogeneity itself changes, providing context-specific insights.

Conclusions:

• Location-scale meta-analysis offers a more holistic approach to interpreting ecological and evolutionary data.
• The framework enhances the biological relevance and practical application of meta-analytic results.
• Enables more informed and context-specific conclusions regarding environmental changes and their impacts.