[Air-L] To blur or not to blur? Twitter profile pics and handles

[Joseph Reagle]

On 06/12/2015 02:35 PM, Jodi Sperber wrote:
> I'm in search other's experience or advice regarding the use of Twitter
> handles and profile images within research. Not a new topic, I know. But
> one that is far from settled.

Hi Jodi, I thought I'd comment on this only because I recently blurred
some images in a presentation about my new book "Reading the Comments."
It was a screen shot of Zuckerberg's Facemash (a hot-or-not-like
predecessor to Facebook using purloined photos from Harvard's student
directories). I found the image online, so the cat was already out of
the bag (for a decade) but I didn't want to condone his actions. That
said, and as you noted, the AoIR ethics guidelines asks lots of
questions [1].

[1]: http://aoir.org/reports/ethics2.pdf

> My dissertation research is focused on a specific health community formed
> on Twitter, bound by a specific hashtag. They host a popular weekly chat,
> but frequently post otherwise as well (overall, they average around 1400
> tweets/week). There is nothing I can see that would be embarrassing or
> private about the pics and handles; to the contrary I think they add
> context and depth to the content as they often provide insight as to how a
> user chooses to present themselves.

In the spirit of [1]:

- what does Twitter's ToS say about using their avatars?
- would the sources themselves be surprised or upset?
- are you treating these sources as public and identifiable; is that how
they are represented in your IRB?
- even if you understand people on Twitter to be public sources (as I
often do) the fact that they are discussing health concerns raises
possible concerns about vulnerability.