Here is the process for using AHA data at the NBER. Our agreement allows us to have up to 100 users of the Annual Survey Database and 10 users of the IT database. PIs must be Research Associates or Faculty Research Fellows of the NBER. Our data begin in 1970.
Any NBER affiliate can apply for access. They would need to fill out and sign the certification and the agreement. Every person touching the AHA data (including and especially student RAs) is considered a user and has to sign the AHA certification form. They are tracked in the DUA database and the corresponding IRB. We should also have an abstract and signed user certifications for each project. Although the AHA is not human subjects data, we send the materials to our IRB so she can reach out to the investigators to inquire whether they are matching to human subjects data. In most cases, they are and we make sure they get the proper IRB review. Once that step is complete, we can set up the data access.
The agreement does not allow users access to the full file, rather the user submits a list of variables and years and NBER IT staff prepare an extract. Currently, the list is in the form of a Stata .do file. See example.do for reference. Available variables differ by year. Many variables change from byte or int to str in more recent years. All variables in the extract will be of type str. Please check that any requested variable is available in at least one of the requested years. That field will be present with a missing value indicator in other years. Note that "long" is a reserved word in Stata, we have renamed the variable "lon". Lastly, note that in the .dta, all variable names are converted to lower case.
Whoever is collecting the materials from the project team can forward to material (preferrably in a batch) to email@example.com for action.
The extract will be made available on our system as aha-extract.dta in your home directory on the non-VPN side of the system. By default only you and members of your group will have access to the file, if you wish another RA or coauthor to have access, tell me and I will put it in their home directory. If you want to copy the data to the aging cluster, log in to a machine on the aging cluster and issue this command:
Note that the AHA-IT data was not collected in 2011 and we did not subscribe in 2014.
Our agreement with AHA requires that the files be used on our servers, and not downloaded. Row or column aggregates may be downloaded, as can small numbers of individual datapoints.