Monthly Archives: February 2014

Real People Silhouettes

I recently had the opportunity to talk to the Church lab about Open Humans, and in the course of preparing slides I needed to represent “people” somehow. The standard “person” icon felt irritatingly male-default (the one on men’s bathroom doors), and mixing it up with the stereotyped “dress equals woman” felt like a miserable propagation of gender binary. So I looked for some silhouettes of people (and I looked for them on OpenClipArt, where all artwork is supposed to be public domain).

I found these1 – and they’re better than icons. People, after all, vary. Humans aren’t a bunch of identical rubber stamps.2

Four silhouettes

You’ll probably think I’m too picky, but I was still dissatisfied. There was only one woman, and she still felt … extra feminized. I mean, it’s not offensive at all, she looks very normal. But the hair, the legs, the shoes… it still felt like a “performance of femininity” and not just “here is a person that happens to be female”.

Recently I saw that Inkscape has a trace tool and thought I could try my hand at making some “real people silhouettes” from real photos. So I searched Flickr and found this CC-BY photo by downstairsdev. From that I made these three figures:

Silhouettes of three people

Can you guess which of these are female? (Answer: all of them.3) Chris wasn’t sure if these were female or male when glancing at them. I think we are collectively inured to the exaggerations made by media: men and women aren’t as different as we make them out to be.

I’d like clip art with less gender-exaggeration to be available to all, so I’ve put them on Wikimedia Commons. I’ve marked these as CC0 because I think clip art really shouldn’t expect stuff like attribution.4 The photo was CC-BY (and not CC-BY-SA) so I think this licensing choice is allowed (let me know if I’m mistaken). (Sorry, I’ve requested deletion of the images per Sage’s correction below.)

I also made a nice “man with a cane” silhouette from this photo by ragesoss of his grandfather (going by the title there) but the photo is CC-BY-SA. As a result I’m not really sure what to do with it I’ve shared it with a CC-BY-SA license, but I really wish it could be CC0. Maybe Sage can change the license to CC-BY…. ;) Thanks to Sage relicensing this, I’m able to share it as CC0.

man_with_cane_silhouette


1: All four were by rejon, links: Person Outline 1, Person Outline 2, Person Outline 3, Person Outline 4
2: Indeed, the fact humans vary — and that their risk preferences vary — is why something like Open Humans should exist. Some participants will be okay with publicly sharing data that others prefer to keep private.
3: If you crank up the curves – or the gamma on your computer – you can see the faces of the silhouetted folks in the original photo.
4: Well, to be fair, I decided to CC0 all my media shared on Wikimedia Commons.

django-study-enrollment

With Open Humans we want to create a site where research participants can share research data so other researchers can re-use and build upon it. One of the first steps here is likely to be creating an enrollment process (similar to what is used in the Harvard Personal Genome Project). Even if we don’t end up using an IRB and consent process, I think the work is useful: right now I know of no turnkey solutions for the general issue of administering an online study enrollment process. Also, at Open Humans I’m sure we’ll want to facilitate online processes researchers use to work with participants.

So I’ve been working on what I hope will be a Django app researchers can use to set up and manage the “study enrollment” aspect of a Django-based research study management site. (It’s admittedly hard to find time to spare.) I also want to have a stand alone site using the app that can be an easy “upload & run study enrollment” software solution with a platform-as-a-service host. A description of the aspirations is also on the current README.md.

We’ve set up an OpenHumans organization on GitHub where we can share open tools like this. I copied my work over to OpenHumans/django_study_enrollment, even though it’s not ready yet, so I have some excuse for an Iron Blogger post. (And maybe to push myself to further improve it. Let me know if you have thoughts, of course.)