Sunday, November 11, 2007

Births 1990 - 2005

The births from 1990 - 1997 are loaded in one RData files and 1998 - 2005 into a second RData file.
The next step is to recode factor values and get a common set of variables. The variable names changed after 2002.

The variable and coding schemes changed after 2002. In order to merge the 1990-2002 data with the 2003-2005 data variables that changed coding need to be modified to have a common coding scheme. The following codes which are used in some but not all years need to be merged into a consistent coding scheme: [Mom|Dad]HispanicOrigin and [Mom|Dad]HispanicCode and [Mom|Dad]Hispanic, [Mom|Dad]EducationCode and [Mom|Dad]Education, [Mom|Dad]PredominantRace and [Mom|Dad]White.

There are still too few births in the other race categories to include them in the analysis. [Mom|Dad]White is coded Yes/No with Yes equal to White and No to Black(African/American).

Coding of education changed from number of years to categories. I have recoded the categories back to years.

Missing values are coded with 9's and will be recoded to NA's