Table 3.
False-Positive Rate of FE, RE, and the New RE at Threshold α = 0.05
| # Studies | Sample Size | FE | RE | New RE |
|---|---|---|---|---|
| 3 | equal | 0.0506 | 0.0381 | 0.0501 |
| 3 | unequal | 0.0504 | 0.0368 | 0.0488 |
| 5 | equal | 0.0493 | 0.0370 | 0.0496 |
| 5 | unequal | 0.0495 | 0.0364 | 0.0490 |
| 10 | equal | 0.0504 | 0.0394 | 0.0503 |
| 10 | unequal | 0.0495 | 0.0375 | 0.0484 |
| 20 | equal | 0.0499 | 0.0406 | 0.0497 |
| 20 | unequal | 0.0496 | 0.0395 | 0.0485 |
A sample size of 1000 is assumed when sample sizes are equal. For unequal sample sizes, we use evenly spaced values such as 100, 200, …, 2000 for 20 studies.