←Back
How to Code School Names More Efficiently: Common Sense, Scripting, and a Novel SAS Application
Sponsor: Survey Research Methods Section
Keywords: SOUNDEX, upcoding, natural language processing, data processing
Akbar Akbari Esfahani
University of California Los Angeles
Matt Jans
UCLA
Ninez Ponce
University of California Los Angeles
Carl Ganz
University of California Los Angeles
For any survey, upcoding is an important, but often time consuming process. At the California Health Interview Survey (CHIS), child and teenage respondents are asked the name of the school they attend. The respondent-reported school names need to match the official names of the schools on the Department of Education (DOE) records for the data to be useful. Here we present an approach for matching school names using the SOUNDEX algorithm, along with a novel SAS application to minimize the amount of time spent manually coding.