In my AP Stats class a learner wanted to see if there was an association between poverty or income and graduation rates in the local Washoe County high schools. This project was done in May and it is an initial project that does have some shortcomings. First, I will explain the methodology the learner used to construct the data, explain the results, and then the shortcomings that can be addressed in a future study.
First, the learner gathered a list of high schools in Washoe County and collected data from the Census Bureau’s “American Community Survey” to gather information about the corresponding zip code. The learner did ignore the schools that select their learners such as Truckee Meadows Community College High School, the Academy of Arts, Computers & Technology and charter schools. Including those schools would skew the results, especially since they are not associated with a specific zone.
The learner then used JMP to create graphs and calculate statistics on the data, focusing on only 2 associations; median income on grad rate and %below poverty on grade rate.
R is –.836 for the poverty, graph and r = .727 for the median income graph.
Grad Rates = 89.03 – 1.055*(Percent Below Poverty Line)
Grad Rates = 52.69 + 0.0003837*(Median Income in dollars)
Both of these regressions had statistically significant slopes, but the interpretation of the slopes makes the case even more apparent. For the poverty graph, as the percent below poverty increases by 1 percent, the graduation rate drops by 1.06 percent. Clearly, poverty is having an outsized impact on graduation rate.
For the median income graph, for every $1000 change in median household income, the graduation rate increased by only .38%. Clearly, the larger impact on graduation rate is the percent below poverty, not the rising median income.
The problem with this initial analysis is that not every zip code in Washoe County is represented. I know my school zone covers 3 different zip codes, while some of the other schools split a zip code between them. As an initial analysis the results are very interesting, and shows that a more detailed study needs to be performed to get a better handle on impact poverty has on Washoe County’s graduation rates.
Why bring this up with such a long explanation? Why post this project in such detail? Because it needs to be seen, especially in light of the recent national and international news / discussion on the impact poverty has on education.
Michael Pershan did a very good analysis of the PISA data and it shows that in the US, poverty matters, a lot.
Secretary Paul Reville also did a take down of the idea that poverty doesn’t matter. The association between poverty and low educational outcomes is well established, but the deniers and the ideologues won’t allow for reality to impact their thinking. Mr. Reville’s statement that,
“Some want to make the absurd argument that the reason low-income youngsters do poorly is that, mysteriously, all the incompetency in our education systems has coincidentally aggregated around low income students.”
is as close as it gets to a perfect description of the state of the political attacks on education.
The problem as I see it is an AP Statistics learner in Reno, NV can figure out the reality of the situation with publically available data. When will we stop denying the obvious and start acting on the good data?