Updating Literacy Estimates in the Oklahoma City Metro

Abstract

Project SCHOLAR (Statistical Consulting Help for Organizational Leaders and Academic Researchers) is an undergraduate student statistical consulting service at the University of Central Oklahoma. SCHOLAR students work under the supervision of faculty mentors from the Department of Mathematics and Statistics on various projects submitted from researchers on campus as well as off campus. This year, a SCHOLAR group was asked by the Oklahoma City Metro Literacy Coalition to update estimates of literacy levels in the Oklahoma City (OKC) Metropolitan Area and find factors related to below basic literacy proficiency. The group started by organizing and cleaning the data. Explanatory variables were collected from the 2010 US Census, and were used to develop a model for predicting the 2003 State Assessment of Adult Literacy estimates. The group performed a univariate analysis for each explanatory variable, using the p-value, residual plots, and QQ-plots to determine which variables were significantly related to literacy. The group performed multiple regression with stepwise selection, using all of the explanatory variables resulting from the univariate analysis after eliminating variables due to collinearity. The group intends to use the model along with the 2015 American Community Survey data to compute updated literacy estimates for every zip code in the OKC metro. The software used to perform this research is SAS and R.