Tuesday, November 26, 2013

Changes to Data Sharing as a Result of PCCF Licensing


We had our first request for PCCF data under the new license agreement.

The faculty member wants to match postal codes to census geography at the dissemination area for research purposes. Simple and straightforward, and certainly within the constraints of the PCCF license as I understand it.

He then anticipates being asked to share the data about the selected dissemination areas with colleagues with whom he is doing the research, who happen to be at American universities. Previously, if it was public data (i.e., not DA level), he could have done this, I think, since the researchers would have been able to retrieve the census data from the web site, and he could simply have provided them with a list of the census geographies (not including postal codes) in which they were interested. At the DA level, however, I believe that he would have to do the analysis (which he could share), but not the DA-level data underlying the analysis.

Three questions:

1. Am I accurately describing the “public” data level situation?

2. Am I accurately describing what is normally necessary with the DA-level data:

a. that he do the analysis and share results, and

b. that the underlying data can’t leave his machine, since his colleagues are not at DLI-member institutions?

3. Has the new license for PCCF has added any wrinkles to this picture?


Any aggregate table available through the DLI falls under the STC open licence regardless of the level of geography. This includes the latest addition of the Census profiles at the DA level.

If the researcher is working from a PUMF then you are correct, they can't share the underlying data with their colleagues (though nothing is stopping them from signing a licence themselves).

In terms of the PCCF, I think this is fine... As long as they don't distribute the tool they should be in business.