3-day window data from a study by John France et al.

A replot of their data whereby the outlier data points were considered to in fact belong to the counts of the three days of high birth counts, the outliers having been due to their inaccurate and unreliable methods of guessing at ovulation.