Confidentiality of the sensitive information provided by participants in a research study has always been an area of concern, especially for qualitative researchers. How do I use information from my participants to tell the narrative of my research, while still ensuring appropriate confidentiality is preserved?
A recent paper entitled "No silver bullet: De-identification still doesn't work" by Arvind Narayanan and Edward W. Felten of Princeton University, thoughtfully reminds us that in this era of "Big Data" and access to large public data sets, it's easier than ever to re-identify participants based on "anonymous" data. Location, details of personal history, and other data that is seemingly not specific to any one person can be collated with public sources to narrow down and identify an individual.
One small way to keep pace is with tools that help ensure confidentiality. At Researchware, we've begun this process by including data masking in HyperRESEARCH's Report Builder.
Clicking "Mask Confidential Data" displays the masking feature, allowing the researcher to automatically hide names or other identifying information in their source material, replacing the information with either pseudonyms or an explicit indication that this is confidential data.
As the Narayanan & Felton article reminds us, "Data privacy is a hard problem." As we continue improving HyperRESEARCH, we'll be looking at data privacy improvements, such as automatically searching your data for information that could potentially be used to breach confidentiality and giving you the option of hiding it. Attentiveness to the new dangers to privacy – on the part of researchers and software developers – will help us all ensure research participants that their personal information will stay private.