How to Optimize Measurement Protocols: An Example of Assessing Measurement Reliability Using Generalizability Theory

Area of Science:

• Physiotherapy
• Measurement Science
• Biostatistics

Background:

• Accurate clinical outcome measurement is crucial in physiotherapy.
• Traditional reliability metrics (e.g., ICC, SEM) assess only one error source.
• Real-world measurements involve multiple sources of error.

Purpose of the Study:

• To introduce generalizability theory (GT) for assessing multiple measurement error sources.
• To demonstrate how GT can identify optimal measurement strategies for clinical outcomes.
• To provide tools for detailed interpretation of measurement error in research and practice.

Main Methods:

• Comparison of classical test theory and generalizability theory (GT).
• Overview of GT terminology and methodology.
• Application of GT to minimize error in measuring knee extension power.

Main Results:

• Generalizability theory (GT) partitions measurement errors into multiple sources.
• GT calculates relative and absolute reliability for any measurement strategy.
• GT enables the identification of optimal measurement strategies to minimize error.

Conclusions:

• GT offers a comprehensive framework for understanding and minimizing measurement error.
• The methodology empowers researchers and clinicians to improve the reliability of clinical outcome measures.
• Optimal measurement strategies derived from GT enhance the validity of physiotherapy assessments.