Final Proposal

For my final project I have chosen option 2, the Data Analysis. The dataset that I have selected to use for my final project is the dataset titled “slave-sales-1775-1865”. The dataset that I have chosen includes both numeric and text data. The numeric data includes information such as the year of slave sales, the age of the slave that was being sold in years, the age of the slave being sold in months, and the appraised value of the slave being sold. The textual data that is inluded describes things such as the state where the sale was taking place, the county in the state, the gender of the slave being sold, any skills that the slave may possess, and any defects that the slave may have. The minimum and maximum ranges for the numeric data in the columns are as follows:

Date of the sale: 1742 –  1865

Age of the slave being sold (in years): 0 – 99

Age of the slave being sold (in months): 0 – 11

Appraised value of the slave: -800 – 525000

The ranges for the descriptive data are as follows:

State of sale (sorted alphabetically): Georgia – Virginia

County of sale (sorted alphabetically): Adams – Williamson

Gender of the slave: Male or Female

Skills the slave may possess (sorted alphabetically): Apprentice – Woodcutter

Defects the slave may possess (sorted alphabetically): Asthmatic – Without Fingers

Each row that the chart contains describes the characteristics of a slave that was sold between the years 1775 and 1865. The amount of information that is given about each slave appears to vary depending on whether or not the slave in questions possessed any skills that would make them more desirable or any defects that could make them less valuable to a potential buyer.


There are a couple of comparisons that can be drawn from this data after analyzing the information in the dataset. The first relationship that I noticed almost immediately is how the appraised value of the slaves changed depending on whether or not they were male or female. It does not appear to be true in all situations, but it seems as though in most cases, slaves that were female were valued at a much lower value than slaves that were listed as males. The next relationship that I noticed between the different columns was that slaves that were listed as having a defect were valued much lower than ones that did not. Conversely, slaves that were listed as possessing some type of skill were valued much higher than slaves that did not. Further, any slave that possessed a skill in tasks that were considered skilled labor, were valued at a higher rate than slaves that possessed skills in domestic labor. Another comparison that I noticed is that the value of slaves in terms of age would resemble a bell curve when put on a graph. This means that newborn slaves and slaves of a young age are valued very little. As slaves get older into there 20’s and 30’s they are valued the most. As they age past this age their value begins to decline again as they enter into old age. One more comparison that I was able to draw from analyzing the data is that the slaves that excelled in skilled labor tasks tended to be male, while the slaves that tended to have more domesticated labor skills tended to be more female.

One thought on “Final Proposal

  • April 11, 2016 at 9:49 PM

    The -800 to 525000 range is odd–I haven’t worked with this dataset before, so I suspect that those are some big typos (since we’re getting this data from another class’s transcription project). You can 1. filter those out like with did with the year range in class or 2. check the spreadsheet and edit the typos.

    Great observation with the correlation between age and value. It may be worth testing out if this is true for different categories–ie, were men with valued skills like blacksmiths and bakers still highly valued late in life, when men with common skills like laborer and field hand were starting to be valued less? Did women’s perceived value peak at a different age than men’s values?

    Give some careful thought to how you’re grouping your skills–how do you know what was more valued? If you group something together based on how you perceive its value, rather than how it was valued by the people who made the records, you may skew your numbers for that category up or down.

Comments are closed.