In this report we will examine and discuss the relationship between the race/ethnicity and poverty levels in the district of "X", specifically focussing on the largest population segments of the district, namely african american students and white students. This will be done by examining histograms, bar charts, pie charts, a radar chart, scatter plot and correlation analysis table. All of these visualisations and tables have been compiled using disaggregated data from the district. It is important to note that poverty level has been determined through the use of a proxy indcator in the form of Free Reduced Price Lunch (FRPL). According to the NCES schools that have a FRLP above 75%, meaning that more than 75% of the students qualify for free reduced price lunch, fall into the poverty category. Whether this is an accurate indicator/proxy is still debated by some. In the final part of the report we wil draw our conlusions and look at why we are seeing what we are seeing.
In chart 1 we can clearly see that there is rougly the same amount of white and african american students in the district. The third largest segement is that of hispanic students, making up 17.32% of the total number of students in the district.
Visualized in another way, the bar chart below (chart 2) clearly shows us the two groups that make up the bulk of the total number of studentsNow let's take a look at the percentage of schools in the district that could be considered "high poverty" according to the NCES. In the pie chart below (chart 3) we can see 43.84% of schools in the dirstrict have FRPL of above 75% thus qualifying them as high poverty.
Read alone chart 3 is already relatively worrying seeing that more than 40% of a districts schools are classified as high poverty. But what becomes really worrying is chart 4 (below) where this information has been grouped by race/ethnicity.
In chart 4 we see that african american students make up nearly 60% of the total number of students in high poverty shools and white students only about 8%. On the other hand white students make up nearly half of the total number of students in shools that are not considered high poverty. This disparity is further visualized by the radar chart below (chart 5) in which the disproportionate relationship between white students in high poverty and non poverty shools is clearly visible.
To further demonstrate this disproportionate relationship between black and white students in high poverty and non poverty schools we can look at the following two histograms (chart 6 and 7).
When we look at chart 6 we can see that most high poverty schools have a total of less than 10% total white students. It is clear that this distibution is seriously skewed. In chart 7 we can essentially see the inverse of chart 6 in that lower percentages of black students in schools makes up the bulk of the non poverty distribution.
Finally we can look at a scatter plot (chart 8) and correlation analysis table (Table 1) to examine the relationship between the percentage of white students per school and the percentage of students who qualify for FRPL (poverty proxy)
What can be see here is a near text book example of negative linear correlation! The r squared value is 0.93, meaning that the linear regression model can explain 93% of variation in the dependant variable. The precise correlation value is - 0.96 meaning that there is a near perfect negative correlation between the two variables. Ultimately, we can say, with a high degree of certainty, that if the percentage of white students is high the degree of poverty will be low and visa versa.
It is important to point out that strong correlation does not necessarily imply causation, especially in this case. The reasons for the relationship seen here are systemic and socio-economic in nature. The correlation can easily be explained by looking at the fact that African Americans (and other minority groups) have historically been descrimainated against on a plethora of different levels. In the US, a history of segregation and intense discrimination has translated into a present day situation where African Americans earn less, live in neighborhoods that are under resourced, have large households with many dependents and, essentially, are more dependent on the state. A further analysis looking at data like georgraphic area, household demographics, annual budget per school and household income would be very interesting and could potentially illustrate what causes the strong negtive relationship between the variables in chart 8.