Draft Story

For my final project I will taking a look at the Slave Sales from 1775 to 1865 data set. Specifically the values of slaves in regards to a skill they possess or a defect they might have. When looking closely at the data set it contains information detailing what you’d expect to find; like age, gender and appraised value of a slave. What I find to be interesting about the data set is it includes skills and defects something I would not have considered. With this information there are more factors that are incorporated into the sale of a particular slave. I believe there are connections regarding the prices of slaves within a particular state possessing a particular skill and which gender the slave happens to be. The females and males tend to have gender specific roles in terms of skills they were labeled with. The men have skills like mechanic, and field laborer and with that their appraised value shows, women have skills such as cooking and baking as well as housework and hairdressing and their appraised value shows as well. While some of the men and women have share skills there are differences that find it harder to compare value based on the skill. However, the defects tend to run along more similar lines. Both men and women share similar defects like, height whether it is too short or too tall, loss of hearing, loss of sight, and any type of sickness including hernias, cancer or just a general label of “sick”. There is also common trends of labeling defects as run away, drunk or fits meaning they are difficult to deal with. From these defective labels you see a shift in appraised value. A drunk man is appraised around $425 where an appraised drunk female has and average value of $300. Maybe not such a shocking revelation seeing as today there is a gender wage gap, however, it became interesting with such defects like, without fingers, a male without fingers is appraised nearly twice as much as woman without fingers, same as a man with cancer compared to a female with cancer.
What I found most interesting or should I say troubling with this data set is the way I read it. At first glance it was troubling to see. These are people and these are children given a price and given a skill or defect and then sold. We all know the horrors of slavery but when you are asked to put it in a bar graph or pie chart you become removed from the fact that these were people and not just numbers on a page. That’s why it’s so important to remember what the numbers really are, and tell the stories of what the numbers are telling us.

2 thoughts on “Draft Story

  • April 20, 2016 at 9:47 PM
    Permalink

    Very good observation in your last paragraph; this is something that’s difficult no matter what dataset you’re working with, but this one especially so. One way of dealing with this in our writing is to use what’s called people-centered language: ie, “plantation owners valued enslaved men’s skilled labor,” rather than “male slaves were valued for their skills,” because the first puts emphasis on the enslaved person as a person rather than as an object. Another way of dealing with is to refer back to the source we’re working with: “in the sales record, this enslaved person was referred to as a drunk,” or “enslaved people described as defective,” rather than just accepting that that characterization of that person is absolutely true. It is a very difficult source to write about.

    Looking through your sheets in more detail than we were able to in office hours, I think you’re in better shape than you thought you were. Some general things to help make your analysis clearer: in the two sheets “State Appraised Value” for skills and defects, I’d suggest switching from your current stacked bars to side-by-side bars (available in show me). Stacked bars are going to total up all appraised values for all states, which right now makes it look like blacksmiths were valued much more on average than any other skill. If you switch it over to side-by-side, you can see that there’s actually a lot of variety within the category blacksmith.

    In the screen shot I attached, I switched the order of skills and State Code in the Columns section–the order you put things in the column section will change how the sections are broken apart in the visual. Put state first, and your bars are grouped by state and then broken apart by skill; put skill first in the columns section and your bars are grouped first by the skill and then broken apart by state. How you order them depends on what you want to emphasize: differences between states, or differences within states? The principle is the same for if you want to group categories by gender, defect, etc etc.

    For the two sheets Male/Female Appraised Value, remember that in office hours we switched Number of Records from dimension to measure (right click), and used the number of records to size the boxes. You could do the same with your first two bubble pack sheets as well.

  • April 20, 2016 at 9:53 PM
    Permalink

    In your sheet State/County/Age/Value scatterplot, I’d suggest switching county code to detail, which will keep your symbols broken apart by county, but using State Code for both color and symbol–this will help you see the variety between counties within a state, but also show more clearly the state groupings. Since you’re interested in gender/occupation differences as well, you might want to add gender into your columns section so you have a male chart and a female chart, and/or using skills instead of state. (When I did a trial run with skills instead of state, there were some pretty clear trends for women but not for men, so take that for what it’s worth).

    Regardless of which direction you decide to take your scatterplot, I’d also suggest putting a filter on both the Average Age Yrs and the Average Appraised Value to filter out the zero values, which pull your averages down. These are ok to filter out because a zero in this case just means no data was recorded.

Comments are closed.