Tracy's sample includes 95 novels by men, and 102 novels by women. As such, when, for example, masking is included in 19 novels by women and 19 novels by men, this ought not be treated as a perfect 50-50 split. For the proportions to actually tie, the sample would need to include slightly more books by women, as in the case of the 28 duels by women and 26 duels by men.

The percentages were weighted as the last step before the creation of each graph. Using Excel to do the calculations, I found the percentage of each gender's books that contained the motif in question, and graphed the relationship between those percentages. This produced results that, generally, added 2% to the 'male' half of the chart.