College Estimation Tool
Alejandro Fraga
Sep 2016
Overview & Purpose
This tool was created to help parents estimate the college cost for their children.
- Uses a time series model to forecast the cost based on kid age
- It integrates an inflation factor for better estimation
- Analyze both private and public college costs in US
The tool was deployed in Shiny under https://enframy.shinyapps.io/College_Cost_Estimation_Tool/
Analysis Methods
The Time series analysis uses the forecast library in R and the HoltWinters for linear data (beta=TRUE)
# Forecast function
forecast_tuition <- HoltWinters(tuition, beta=TRUE, gamma = FALSE)
# Forecast number of years
lastattendance_yr <- remyears+3
forecast_tuition_future <- forecast.HoltWinters(forecast_tuition, h=lastattendance_yr)
Sample Output
As output creates a plot showcasing the trend and future costs
References
The data set used to analyze the linear model was gathered from:
The College Board, Annual Survey of Colleges; NCES, IPEDS data.
- 1987-88 and after: data from Annual Survey of Colleges, the College Board, weighted by full-time undergraduate enrollment;
- 1986-87 and prior: data from Integrated Postsecondary Education Data System (IPEDS), U.S. Department of Education, National Center for Education Statistics, weighted by full-time equivalent enrollment