Target Validity and the Hierarchy of Study Designs

Area of Science:

• Epidemiology
• Biostatistics
• Health Research Methods

Background:

• External validity, encompassing generalizability and transportability, is increasingly critical in research.
• Current methods often assess internal and external validity independently, potentially limiting comprehensive evaluation.
• A unified approach is needed to address the limitations of isolated validity assessments.

Purpose of the Study:

• To propose a novel measure, target validity, that integrates internal and external validity.
• To formally define target bias and target validity for assessing effect estimates in specific populations.
• To illustrate the utility of target validity in comparing research findings.

Main Methods:

• Formal definition of target bias as the discrepancy between true causal effects in a target population and estimated effects in a study sample.
• Definition of target validity as the condition where target bias equals zero.
• Application of the target validity concept to diverse research scenarios and comparisons.

Main Results:

• Target validity provides a unified framework for assessing the validity of effect estimates.
• Even perfect internal validity does not preclude bias in a specific target population.
• The proposed measure clarifies comparisons between experimental and nonexperimental research.

Conclusions:

• Considering internal and external validity separately can be problematic for research interpretation.
• Target validity offers a more comprehensive approach to evaluating the applicability of study findings to specific populations.
• This framework enhances the understanding of causal effect estimation across different research contexts.