IALSA / HRS

Shaping data from the Health and Retirement Study.
GNU General Public License v2.0
5 stars 2 forks source link

Data cleaning for number of close children #12

Open casslbrown opened 7 years ago

casslbrown commented 7 years ago

In 2004 the maximum of closechild is 13. In 2006 the maximum was 66, In 2008 the max is 99, then 66 again in 2010, then 80 in 2012, then 99 in 2014. In the codebook for some years it states that 98 is DK; or NA (Not Ascertained) and 99 is Refused. In some years of the codebook the value 66 is included as code for an "out of range" value.

Data cleaning rules:

  1. Recode values of 66, 99 or 98 to NA
  2. If the number of close children listed was a double digit (e.g., 22, 33, 44) the number of children was made equal to the single digit. [This solves the problem for the majority of cases with greater than 20 close children from 233 to 86]
  3. Otherwise recode the value to NA.