Collaborative Fall 2021 Data Science project
By: Nick Ouellet and Garrett Gilliom
Live webpage found here.
This repository stores the source code and data used for a data-based analysis tutorial of K-12 Academic Performance of Louisiana and New Jersey school districts.
The team investigated whether financial and enrollment variables held any relationship with academic performance metrics in both states. Initial findings suggested that little to no relationship existed between features such as income/expenses/size of schools and graduation rate/ACT scores – which contradicted the team's initial expectations.
The tutorial concludes with the creation of a model based on New Jersey data that attempts to predict the graduation rate of Louisiana school districts. Although the created model consistently predicted average graduation rates that were higher than those in reality, the team is cautious that any true takeaway should be made from these results, considering the original EDA displayed little relationship between the features (finances, enrollment data) used and the label (graduation rate) predicted.
If a relationship between the variabes was found and these same results occurred, then it might serve as evidence of some other variable or quality of the systems that is causing a difference between the predicted and true values. However, even if this was the case, the team stresses and recognizes that this analysis is limited in scope by both time and diversity of schools; a single instance of a relationship, or the absence of one, in this case, is not indicative of the entire system's qualities. As such, the team recommends that more research be conducted to determine whether relationships between these features of school districts exists by looking into more states and timelines.