Even more on gender bias in textbooks. Refer to the previous exercise. Let's now combine the categories "girl" with "woman" and "boy" with "man."
For each text calculate the proportion of male references and test the hypothesis that male and female references are equally likely (that is, the proportion of male references is equal to 0.5).
Summarize the results of your 10 tests. Is there a pattern that suggests a relation with the gender of the author?