John Verzani

Using NHANES III data to create growth charts

The CDC publishes growth charts based on NHANES III data at http://www.cdc.gov/GrowthCharts/. In addition to the charts, there are some educational materials (including SAS code which I haven't looked at). This project would demonstrate to students:

  1. various skills needed for working with large, government-style, data sets, including: working with data in fwf format, data cleaning, possibly using a data base for storage. (http://www.cdc.gov/nchs/nhanes/nh3data.htm#NHANES%20III%C2%A0%20Series%2011,%20No.%201 )
  2. Allow students to look at various rubrics to predict adult height such as the doubling of height at 2 years.
  3. explore model building for this data. This data shows how the simple linear model is not sufficient, how using polynomial models can fit the data, and how modeling of the quantiles is more natural. For this, the quantreg package is available. The paper http://www.econ.uiuc.edu/~roger/research/growth/growth.html gives background material for the use in this context.