Statistical power analysis
To compare statistical power and effect size between our replication and
the original study, we calculated these parameters using the “pwr”
package (Champely, 2020). With a sample size n = 20 and alpha = 0.05,
Samaha & Postle’s reported correlation coefficients of r = -0.503 (for
eyes-closed data) and r = -0.605 (for pre-stimulus data) resulted in a
statistical power of 0.644 and 0.84, respectively. With our sample size
of n = 31 and alpha = 0.05, correlations in our replication would need
to be r = 0.41 or greater for eyes-closed data, and r = 0.5 or greater
for pre-stimulus data, to match or exceed the statistical power of
Samaha & Postle’s original study.