Sign in or register
for additional privileges

The Hannah More Project

Computational Analysis, Author Attribution, and the Cheap Repository Tracts of the 18th Century

Cluster Analysis, page 1 of 1
Previous page on path 

 

You appear to be using an older verion of Internet Explorer. For the best experience please upgrade your IE version or switch to a another web browser.

Cluster Analysis Results

On this page are the visualizations generated by the Cluster Analysis team.
 
To view the visualizations: Clicking on the title of each visualization in the descriptions below will scroll to the appropriate thumbnail on the right. To see a larger view of the image, click the "Details" button on the bottom right corner of the thumbnail.

Combination Grouping (of Hannah More and not signed texts):


Delta Cluster Analysis

A dendrogram showing the similarities and differences between Hannah More texts (represented in gray) and the not signed texts (represented in different colors). Here, "Charles Jones" and "The History of Mary Wood" pop out among the Hannah More texts, thus suggesting that Hannah More authored these texts. 

Bootstrap Consensus Tree: 

A Bootstrap Consensus Tree graph showing the more subtle groupings between the two larger data sets. There is a clear branching off of the texts in the subgroups that highlight commonalities within Hannah More and not signed texts. For example, the following Hannah More texts share similar textual patterns: On Envy, On True and False Meekness, Thoughts on Conversation, Heart & Temper in the Education on Daughters, Importance of Religion to the Female Character, On Dissipation, On Danger of Sentimental or Religious Connexions, and Miscellaneous Observations on Genius.

PCA Label Graph: 

A splatter graph showing a large corpus of texts (and their titles) that share patterns in common and, more interestingly, also show outlying texts. These outlying texts ("The Gamester," "John the Shopkeeper Turned Sailor, pt. II," and "The History of Mary Wood") though they might be grouped with other texts in the dendrogram and the Bootstrap Consensus Tree are separated from the other texts in this representation showing that they have elements of individuality which is also taken into consideration in our final analysis on author attribution.

PCA Dot Graph: 

A dot splatter graph showing the same results as the splatter graph with text titles but here the overlapping of individual texts are clearer. For example, the overlapping red dot and black dots near the bottom of the graph correspond to the black and red titles in the previous splatter graph. This suggests that "The Christmas Story" and "The Second Christmas Story" were written by the same author. 

Hannah More Grouping of Texts:


Delta Cluster Analysis: 

A dendrogram showing the similarities and differences within the Hannah More text sample. This allows a more detailed analysis of the Hannah More tracts and gives more context for the combination dendrogram results. For example, "The History of Tom White" and "A Cure for Melancholy" are paired off showing that they are very closely related in word usage, which shows that Hannah More (knowingly or unknowingly) used the same textual framework when writing on certain themes.

Bootstrap Consensus Tree: 

A Bootstrap Consensus Tree showing the branching off of specific text groupings in the Hannah More tract sample. For example, the grouping of "Tawney Rachel," "Betty Brown," "The History of Hester Wilmot," "The History of the Widow," and "Black Giles and the Poacher" show common textual patterns and magnify the results for Hannah More tracts that were viewed alongside the larger combination Bootstrap Consensus Tree. 

PCA Label Graph: 

A splatter graph showing outliers in the Hannah More tract sample. Specifically, "A New Christmas Hymn" and "Address to Persons Attending a Funeral" are outliers on the graph, which suggests that either these texts weren’t authored by Hannah More or that the textual patterns differ due to a change in the author’s style because of the topic or another unknown reason.

PCA Dot Graph: 

A dot splatter graph showing the same results as the splatter graph with text titles but here the overlapping of individual texts are clearer. For example, the overlapping purple and tan dots near the top of the graph correspond to the purple and tan titles in the previous splatter graph. This shows that "Black Giles and the Poacher" and "Betty Brown" have very similar textual patterns and suggests that Hannah More scholars should pay close attention to tracts that follow this pattern to elucidate on the reasoning behind these similarities.

Not Signed Grouping of Texts:


Delta Cluster Analysis: 

A dendrogram showing the similarities and differences within the not signed tracts text sample. This allows a more detailed analysis of the not signed tracts and allows us to better group off the text sample. For example, "Some New Thoughts for the New Year," "Onesimus," "Hints to All Ranks of People," and "The Death of Christ" are very closely grouped together and suggest a common author. 

Bootstrap Consensus Tree: 

A Bootstrap Consensus Tree showing the branching off of specific text groupings in the not signed tract sample. For example, the grouping of "The General Resurrection," "The Fall of Adam," "Joseph and His Brethren, pt. I," and "Joseph and His Brethren, pt. III" show common textual patterns.

PCA Label Graph: 

A splatter graph showing outliers in the not signed tract sample. Specifically, "The Gravestone," "The Hampshire Tragedy," and "John the Shopkeeper Turned Sailor, pt. II" are outliers on the graph.

PCA Dot Graph: 

A dot splatter graph showing the same results as the splatter graph with text titles but here the overlapping of individual texts are clearer. For example, the overlapping purple and mauve dots in the center of the graph correspond to the purple and mauve titles in the previous splatter graph. This shows that "Wild Robert" and "King Dionysius and Damocles" have very similar textual patterns and suggests that these two texts were very likely authored by the same person.

Comment on this page
 

Discussion of "Cluster Analysis Results"

Add your voice to this discussion.

Checking your signed in status ...

Previous page on path Cluster Analysis, page 1 of 1