In a repeated-measures design, how does the design typically affect error variance compared to a between-subjects design?

• A. Error variance increases due to using the same participants.
• B. Error variance is typically reduced because participant differences are controlled.
• C. Error variance is unaffected by design.
• D. It eliminates error variance entirely.

Answer: (B)

The main idea is that a repeated-measures design reduces error variance by using the same participants in every condition, so differences between people don’t inflate the variability inside conditions. Since each person serves as their own control, the stable, individual baseline tends to cancel out when comparing conditions, making the remaining unexplained variability smaller. This tighter, within-person comparison increases statistical power to detect effects because the error term is reduced relative to a between-subjects design, where different people bring their own distinct baselines and traits. Of course, it doesn’t eliminate all error—there’s still measurement error and potential issues like practice or fatigue effects and carryover—but the dominant effect is a reduction in error variance due to controlling for participant differences.