Figure 5
The diagram is titled “Distribution of Token Counts by Category (Top 15 Categories)”. The vertical axis is labeled “Token Count” and ranges from 0 to 10000 with increments of 2000. The horizontal axis lists fifteen categories: “Literary underscore forgeries”, “Pseudepigraphy”, “Old underscore Testament underscore pseudepigrapha”, “Forgery underscore controversies”, “Archaeological underscore forgeries”, “Musical underscore hoaxes”, “Art underscore forgeries”, “Document underscore forgeries”, “Ancient underscore Greek underscore pseudepigrapha”, “Political underscore forgery”, “Religious underscore hoaxes”, “Modern underscore pseudepigrapha”, “Sculpture underscore forgeries”, “Anti underscore Islamic underscore forgeries”, and “Shakespeare underscore authorship underscore question”. Each category is represented by a vertical boxplot showing the distribution of token counts. The box indicates the interquartile range, the horizontal line inside each box represents the median, vertical whiskers extend to non-outlier minimum and maximum values, and circular markers above the whiskers represent outliers. For “Literary underscore forgeries”, the median appears below 1000 tokens, with outliers extending near 8000 tokens. “Pseudepigraphy” shows a median below 1000 tokens, with outliers near 4500 tokens. “Old underscore Testament underscore pseudepigrapha” has a median under 1000 tokens, with upper values approaching 4000 tokens. “Forgery underscore controversies” shows a median near 1000 tokens, with outliers above 7000 tokens. “Archaeological underscore forgeries” has a median near 1000 tokens and upper values exceeding 6000 tokens. “Musical underscore hoaxes” shows a lower median near 500 tokens and fewer extreme outliers. “Art underscore forgeries” has a median below 1000 tokens with several outliers above 5000 tokens. “Document underscore forgeries” shows a median under 1000 tokens and outliers exceeding 6000 tokens. “Ancient underscore Greek underscore pseudepigrapha” has a median below 500 tokens and a moderate spread. “Political underscore forgery” shows a median near 1000 tokens with outliers above 8000 tokens. “Religious underscore hoaxes” displays a wide spread with a median below 500 tokens and whiskers extending above 6000 tokens. “Modern underscore pseudepigrapha” has a median below 1000 tokens with moderate spread. “Sculpture underscore forgeries” shows a median near 1000 tokens and moderate variability. “Anti underscore Islamic underscore forgeries” has a median near 1000 tokens with a narrower spread. “Shakespeare underscore authorship underscore question” shows the highest median among the categories, above 2000 tokens, with upper values exceeding 3000 tokens.Token count distribution by category showing medians, quartiles and outliers. Source: Authors' own work