As an RDC analyst, I would like to share with you some advice I give to all researchers whose proposals are assigned to me.
As you know, it is a researchers’ responsibility to extract and send public-use NCHS and non-NCHS data to the RDC to be merged with restricted variables by their analyst. It is recommended that you familiarize yourselves with NCHS data by doing preliminary analysis with the public use data. Often you may need to rename, recode or re-categorize variables for your analysis. It may seem like a good idea to send us a public use dataset with derived variables instead of the original ones. I strongly urge you not to do this.
While it is not against RDC rules to send us recodes instead of original variables, doing so may lead to extra work for your RDC analyst as well as delays and extra charges. There are two examples that come to mind. On both proposals, researchers sent in derived variables instead of the original ones. Researchers working on the first proposal made a mistake while creating derived variables. With regard to the second proposal, the Student advisor changed her mind about the grouping of analytic variables and the researchers needed to categorize them in a different way. Since the original variables were not included in the data sets the researchers sent to the RDC, the researchers had to resend the public use datasets with original variables and I had to redo the merge. This resulted in delays for both projects and additional data setup fees.
The conclusion is: send your analyst the original variables, instead of the derived ones! Following this simple rule will save RDC analysts time as well as your time and money. If you want to create derived variables and keep them on your permanent analytic dataset, just send us your programming code to create such variables. Your RDC analyst can either run the code while creating your analytic dataset or put your code into your folder along with the data so that you can create the derived variables yourselves.
Signed RDC Analyst