Adding subjects or adding measurements: Which increases the precision of longitudinal research?

Area of Science:

• Biostatistics
• Clinical Research Design
• Longitudinal Studies

Background:

• Designing repeated measurement studies requires balancing the number of subjects and measurement times.
• Researchers often question whether adding subjects or measurements yields greater statistical precision.

Purpose of the Study:

• To present a method for evaluating the relative benefits of increasing subjects versus measurement times in repeated measures studies.
• To determine the optimal number of follow-up assessments for reliable change detection.

Main Methods:

• Utilized the standard error of estimate (SE) for mean change as the primary precision criterion.
• Analyzed an existing dataset of post-stroke patients with six follow-up assessments across six rating scales.
• Calculated and compared SE values for two common change indices across various combinations of measurements (2-6).

Main Results:

• Presented sample sizes needed to achieve the same precision gain as adding an extra measurement.
• Demonstrated that five or six repeated measurements appear sufficient for accurate change assessment.
• Indicated diminishing returns in precision with more than six measurements.

Conclusions:

• For longitudinal studies, five to six repeated measurements provide adequate precision for assessing change.
• To enhance statistical precision further, increasing the number of subjects is more effective than adding more measurement occasions.
• The findings offer guidance for optimizing resource allocation in repeated measures study design.