Project Puppy: getting data clearance

February 8, 2012 – 9:18 pm

As our adventures in radically transparent engineering education research continue, ampoule we find ourselves staring at IRB and going “well, all right, how is that going to work?”

Well, here’s how it’s going to work.

  1. Linda gets individual permissions to make transcripts & identities public (and keeps track of the emails granting permission).
  2. Some participants may request certain edits before their transcripts are released. We do this editing.
  3. Transcripts are posted online under a creative commons license. We may want to post permission emails (stripped of email addresses) alongside the data as the closest thing we have to “signed informed consent forms.”
  4. The Purdue group visits Purdue’s IRB to get these interviews cleared as a public use dataset. Specifically, we’re in category 6b on page 3 of that document.

Or at least that’s the plan as of now. If we need to adjust, we’ll adjust. But if this all works, we’ll have an open data set that can be used by any future Purdue researchers for any future research without any further IRB approval – it basically places it in the same league as, say, publicly available census data.

Of course, this doesn’t solve the problem for other institutions. But Purdue has a pretty good reputation as a research university, so if we post the notice that Purdue’s IRB is ok with it, it makes it easier for other schools to go “well, if Purdue says it’s ok, we guess it’s ok here too.”

Know someone who'd appreciate this post?
  • Print
  • Facebook
  • Twitter
  • Google Bookmarks
  • email
  1. One Response to “Project Puppy: getting data clearance”

  2. Ah – it seems so easy :)

    Below is an overview from a longer conversation that shows just how complicated it is to move in this direction and think of IRB issues as “everyday issues” (meaning, with or without including IRB in our process this is our pledge to follow ethical practices).

    Big questions: What does IRB mean if we do a creative commons license? How can we gauge the ethical treatment of human subjects? Ensure all are protected? What does it mean to ask for IRB approval after data was collected (the original study was a part of a personal journey and was not intended to be publicly shared)?

    Option 1: Write up an IRB application where we retroactively ask for informed consent to make data-interviews of people talking about change-public. This launches into a whole discussion of what to we mean by “public” (us? Our group? People we invite? People “out there in the world” who find our website?). We have to think this through – since having this kind of data public might be risky for some folks – and it is a little different from “triangulation“ (which is part of a qualitative research process in which you go back to the original subjects with a copy of their data and your interpretations of the data and ask for feedback and in some sense validation or an alternative take on the researcher’s interpretation). We also need to think through all the issues around sharing data publicly – for example, what if this encouraged someone who saw the public data to contact one of the subjects? What if someone did their own analysis on the data and published it – how do we protect the rights of the people in the study who are now being represented by others who have not “signed onto the process”? We know there are ways to think about addressing these, but we really need to think through this before we fully go public. Also, typically when we ask for informed consent we have put in place strategies to protect confidentiality (e.g., remove identifiers, limit access to the data) – in this situation there is something important about including identifiers. What kinds of protections are needed in this case?

    Option 2: We convert the data into public data via something like a Creative Commons license. In this case, it is outside of IRB (since it is publicly available data) – but the issues of Option 1 don’t go away – what do protections look like in this case? The big distinction is that the choice about what is public lies with the participant – and are owners of the creative commons license (okay, we’re not a 100% sure about who owns the copyright – the participant or the person who collected the data and had it transcribed). FYI – we have already heard from a couple of people that they would prefer the data gets “cleaned” a bit – e.g., cleaning out extraneous language like “um” (discourse analysis researchers turning over in grave…), having data look more like sentences vs. long incomplete sentences?

    What does this really mean? Ownership of the data or copyright would essentially mean decision making power about what may be used, remixed, shared, etc. All of these issues would need to be communicated in the license itself, including any institutional boundaries regarding the license that might need to be part of an IRB-waiver application (e.g., only Purdue researchers vs. any researcher at any institution). Which means, that this doesn’t make the IRB process go away, it’s just “different” – because as Purdue researchers we are subject to the IRB process at Purdue.

    Option 3: Either of the options above but with the limitation that we do remove identifiers. This is the discussion that seems to be first on the table – that we need to ask the participants what protections they might want. When Linda asked, it seemed like most were open to sharing the data publicly, although some were concerned about how the information might impact them in their current professional role(s) – not that the information was sensitive – but that the act of sharing something publicly means that the information could be used in unexpected ways that might place them (or the projects they work on) at risk.

    Bottom line – this is complex – this requires more discussion including discussion with our participants. This is a partnership. A reason we are pursuing this is because we believe that there is something important about making this public – particularly given the topic of “change”. We need to keep this conversation on-going as we work through the complexities…

    By Robin on Feb 14, 2012

What do you think?