In the SRM calculation, the standard deviation used in the denominator is taken from which source?

In the SRM calculation, the standard deviation used in the denominator is taken from which source?

• A. Change scores
• B. Initial scores
• C. Post-treatment scores
• D. Pooled from baseline and follow-up

Answer: (B)

SRM measures how large the average change is relative to how much change varies among individuals. The denominator uses the standard deviation of the change scores themselves—the differences between post-treatment and baseline for each person. This choice anchors the metric to the actual amount of change each person experiences, so the SRM reflects responsiveness in terms of how consistently people respond. Using the change scores’ variability matters because it directly represents dispersion in response. If everyone improves by about the same amount, the SD of change is small and the SRM becomes large, signaling strong, consistent responsiveness. If changes vary widely—from little improvement to large gains—the SD of change is large and the SRM is smaller, indicating less consistent responsiveness. In contrast, the variability of initial scores tells you about starting differences, not how much individuals changed. Post-treatment scores or a pooled baseline-and-follow-up SD would describe end values or overall dispersion, not the magnitude and consistency of change itself.

SRM measures how large the average change is relative to how much change varies among individuals. The denominator uses the standard deviation of the change scores themselves—the differences between post-treatment and baseline for each person. This choice anchors the metric to the actual amount of change each person experiences, so the SRM reflects responsiveness in terms of how consistently people respond.

Using the change scores’ variability matters because it directly represents dispersion in response. If everyone improves by about the same amount, the SD of change is small and the SRM becomes large, signaling strong, consistent responsiveness. If changes vary widely—from little improvement to large gains—the SD of change is large and the SRM is smaller, indicating less consistent responsiveness.

In contrast, the variability of initial scores tells you about starting differences, not how much individuals changed. Post-treatment scores or a pooled baseline-and-follow-up SD would describe end values or overall dispersion, not the magnitude and consistency of change itself.