With many data points per condition, on average the effect of the noise “washes out”.

1. The average score is close to the true value of the dependent variable (as though there were no noise).

2. The more noise in the data, the more data points you need to get a reliable result.

Author: David Krider