Category: Reflection Blog

I will be focusing on the major buildings here at Bates College. I am interested in looking at where each of the donors earned the majority of their money. I would like to be able to create a visualization for each of the donors that tells a brief story of how they had such extreme financial stability that they would be able to donate enough money for a building. As I have been thinking more about this project the more I am thinking about the understanding that all money has been made because of the inhumane exploitation of enslaved black people. 

I was looking at the Bates website and noticed how boring and uninformative the page on the history of each building truly is. There is a muddled paragraph or two about the finances involved and that is pretty much all. In theory, data visualizations could be used in a really awesome way to best accurately represent the history of the people whose names are commonplace around the Bates College campus. I am expecting there to be some limitations in the amount of information I am able to use to best describe each donor and their respective buildings but I am hopeful. I am not sure if I will do anymore analysis than descriptives about each building but I may attempt to summarize and categorize where the funding of the majority of buildings at Bates College was from. 

For the final project, I would be interested in looking at the Maine State Seminary Data. I believe this was mentioned in class but I think it would be interesting to look at donation amount and distance from Boston. Another interesting thing could be the distance from a major healthcare facility. I think incorporating occupation or literacy rates into this analysis would also be interesting. The results of this analysis would tell us about resource accessibility and donations. 

I would also be excited to look at the cotton data as well, but for some reason, I struggle to conceptualize and visualize that data. I think that both sets of data are super interesting, but with the Maine State Seminary I feel like each row of data is also a person since we have so much more information. I think I prefer the proximity to people that the MSS data offers. The cotton weights data is more numerical which feels less realistic. This is likely a reminder that I am a humanities person, not particularly drawn to a page filled with numerical data. 

This code calculates the correlation between the number of bales and total weight and the number of bales and cost per pound. This analysis is conducted by using the is.na() function to identify the location of the missing values (NA’s) and assign it to a new data frame that do not contain the missing values. This process is done for each of the columns which are Number of Bales, total weight, and cost per pound, gradually reducing the data frame to one that contains numeric values in all value locations. Lastly, the number of rows is returned to compare the reduction in rows as we rid the data of the NA values. 

With all numeric values, the new data frame can be used to assess the relationship between each of the columns. This is done using the plot() function and cor() function, one significant relationship is observed in the relationship between a number of bales and total weight. The scatter plot created with the plot function shows a significant positive slope. The correlation calculated additionally determines there is an eighty-one percent chance that the relationship did not occur by chance.This means that the increase in the number of bales likely will indicate an increase in the total weight.

A correlation analysis was conducted on the relationship between cost per pound and the number of bales using the same functions as above. The relationship, in this case, is weak and negative. The analysis found a twenty-six percent likelihood that the relationship occurred by chance. This indicates a fairly insignificant relationship between cost per pound and the number of bales.

I have been really enjoying this class so far. I have loved the combination of data analysis within a humanities context. I am feeling like I have fallen behind a little bit in this class while trying to keep up with everything else that I am trying to get done. I think that I am struggling a bit to understand the code we are doing as it gets more complicated. A pattern I have been noticing is that I will only fully understand the material from the previous lab, once we have already moved forward to working on new material. This means for me right now, that I am struggling with loops. I feel as though I have a good understanding of what the structure of the for loop should look like, however, no matter what I do, I seem to be missing some detail since I consistently am coming up with errors, specifically in Lab 6. I am planning to reach out to Professor Shrout to meet one on one or with my lab partner to discuss the details of this lab and Lab 5 as we fell short in completing Lab 5 as well. 

As we move forward, I am optimistic that I will be able to catch up and keep up with the material. I think these days I have been feeling like I am constantly one step behind and have been stuck in that cycle. I am really excited for the remainder of this class and wish that I had taken this class sooner so I would have a chance to take another course with Professor Shrout.

This visualization includes bins of 5000 years which caused the histogram to be extremely wide scope. This demonstrates that the majority of the data is condensed at much lower donation amounts, however, the outliers are incredibly far. The few donations that are so much money will cause the distribution to be incredibly skewed. The median of the data would be immensely smaller than the mean of the data. These outliers will result in the data having an extremely positive skew making it quite far from normally distributed. 

I chose this binning to investigate the data from a birds-eye view. Massive donations result in the data being difficult to analyze as the mean is not highly representative of the data overall. Within our lab, we showed the data with smaller bin sizes to show the high amount of variability within the data.

# this is the same histogram as above but with different sized bins that makes it more difficult to see each individual bin but easier to see the total overview of the data. 

hist(totalDonations, breaks = c(5000))

Welch Two Sample t-test

data:  donations1857 and donations1854

t = -1.4185, df = 99.361, p-value = 0.1592

alternative hypothesis: true difference in means is not equal to 0

95 percent confidence interval:

 -19.667416   3.269478

sample estimates:

mean of x mean of y 

 12.96986  21.16883 

This is a comparison of the donation amounts received by Bates College in the years 1857 and 1854. The donation amounts from each year are taken from the full set of data and compared to each other. An independent samples t-test was conducted to check the difference between donation amounts in 1857 and donation amounts in the year 1854, t(99.36) = -1.42, p-value = 0.16. No significant difference not occurring by chance was found with the alpha = .01 and donations1857 M = 12.97 and donations1854 M = 21.17. The independent samples t-test we completed tells us how likely it is that the relationship we see between donations from those two years occurred by chance. The results from this test show no significant difference between the donations from those two years which we know by the higher p-value of the test output. 

The context under which one may need to understand the cotton weight data is that Benjamin Bates donated a large sum of money to Owen Cheney, founder of the college, to support the institution. He gave so much money that they named the school after him. Bates made all of this money off of selling textiles. The cotton weight data shows how much money Benjamin Bates made off of selling textiles that were made from cotton picked by enslaved people. The school was always accepting students of color and coined themselves as “founded by abolitionists”. Yet the money that built the school and whose the man who donated it was wealthy off of the backs of enslaved people in the southern United States. The process of collecting this data would be to look at the location of purchase and whether or not the product was from enslavers cotton farms. If the majority of the money comes from such farms, the school is somewhat ignorantly claiming to be anti-racist from the start of the school. The data could encourage the administration to think critically about their somewhat false advertisement.

Thus far, I have been greatly enjoying this class. I am feeling really excited to explore data from a humanitarian perspective. I have loved the intentional focus of Bates College as it offers proximity to students that are essential to encouraging investment in the things we are learning. I also think it is really important to think critically about the institution we are pouring our money into and how we can leave it better than when we started. Better is ambiguous but in this context, it is a safer, happier, more supportive, and welcoming space for everyone to enjoy. That is the aspect of this class that I am most looking forward to. 

I am nervous about the coding aspects of this course as I don’t have a lot of experience with this type of work and I am not excellent with formulaic work. However, I do enjoy the immediate feedback that working with code offers. That is something that is helpful for me to stay invested and focused on my work. I am excited to learn how to work with data in code and to expand my knowledge of how programming works. I am also currently working on a thesis in psychology which will require me to work with lots of data. I am hoping this course will complement my thesis work and help me complete the best data analysis I can do.