Case Study

Apply the sample size benchmarks for an independent-samples tt-test with a large population effect to Dr. Aris's study. Determine how many participants she must recruit per group to achieve a power of 0.800.80, how many she would need to achieve a power of 0.990.99, and justify whether she should adjust her initial plan of 2020 participants per group.

Case context: Dr. Aris is designing an experimental study comparing a new mindfulness technique to an active control group using an independent-samples tt-test. Based on past literature, she expects a large population effect size. She initially plans to recruit 2020 participants per group, but a senior colleague advises her that this sample size might be insufficient to achieve acceptable statistical power.

Question: Apply the sample size benchmarks for an independent-samples tt-test with a large population effect to Dr. Aris's study. Determine how many participants she must recruit per group to achieve a power of 0.800.80, how many she would need to achieve a power of 0.990.99, and justify whether she should adjust her initial plan of 2020 participants per group.

Sample answer: Dr. Aris must adjust her initial plan of 2020 participants per group because it is insufficient to achieve the standard statistical power of 0.800.80. For an independent-samples tt-test with a large population effect, she needs approximately 2626 participants per sample to reach a power of 0.800.80. If she wishes to achieve a very high statistical power of 0.990.99, she would need to recruit 5959 participants per sample. Therefore, she should increase her sample size to at least 2626 participants per group, or up to 5959 per group if she wants to ensure an extremely low probability of a Type II error.

Key points:

  • Identifies that Dr. Aris's initial plan of 2020 participants per group is insufficient to reach the standard power of 0.800.80.
  • Applies the benchmark of approximately 2626 participants per sample to achieve a power of 0.800.80.
  • Applies the benchmark of approximately 5959 participants per sample to achieve a power of 0.990.99.
  • Justifies the recruitment adjustment by linking sample sizes to the desired power levels and control of statistical error.

Rubric: The student must apply the sample size benchmarks of 2626 (for 0.800.80 power) and 5959 (for 0.990.99 power) to Dr. Aris's case. They must correctly identify that her current plan of 2020 is too small to reach even 0.800.80 power, and recommend adjusting the recruitment targets upward to either 2626 or 5959 participants per group with appropriate statistical justification.

0

1

Updated 2026-05-27

Contributors are:

Who are from:

Tags

KPU

Research Methods in Psychology - 4th American Edition @ KPU

Related