For Experimonth: Race, we have a confessional, where people can write about something they don’t feel comfortable posting on Facebook and we have a closed group where those confessions go and experimonthers can talk about them.  

So far there have been over 7,000 words written and I took a moment this morning to visualize the top 100 in this word cloud.

Sorry for the long-time-no-post, but here’s a new thing.

We’ve created an Experimonth Headquarters space at the museum to see how a project that was born digital fares on the floor of the museum.  So far, we’ve realized it in a number of ways, one of which is this skin tone visualization.  Visitors pick paint swatches, match their skin tone as best they can, and slap it up on a sticky wall.  This visualization was from this past weekend.  

I just wrote a post over at Experimonth: Mood about participant compliance that I think folks who follow useum may be interested in.  

In short, we retained 86% of our participants throughout the experimonth and they were 81% compliant with texting in data.  

Lots more info and charts on the post proper.

Experimonth: Mood is trucking along and I thought in celebration of our 10,000th mood rating data point collected, I’d let folks over here at Useum know how things have progressed.

Technical difficulties were rife in the beginning.  I’ve learned a lot about SMS and the infrastructure that supports our ability to send and receive the 300 text messages for each participant over the course of a month.  I still don’t really understand how everything’s connected, but I do know enough to do some things differently next time (like getting a short code so that carriers don’t reject our text messages due to spam filters).

This blog is definitely a team effort.  We have three co-hosts for the Experimonth: myself + the researcher + an exhibit developer and a web developer who’s handling all the bugs and feature requests.  We’re still working out the kinks in getting posts reviewed for scientific accuracy (something I’m not used to in the kind of writing I typically do), and no combination of project management software (we’re currently shuffling between Flow, Dropbox and Google Docs) + email + in person meetings has proven to fit just right yet.  

Traffic to the site has decreased as the month’s gone on, though mood rating compliancy is till through the roof at 80%+.  I’ve got some new graphs and features planned that I hope will re-energize participants to visit the site throughout the latter two weeks.

My favorite part of the month so far is reading comments from the participants.  I’m floored at how forthright folks are about their own life and experimonth experience.  I hope we get a good turnout at our wrap-up event so that I can meet these creative and thoughtful people in person.  

More soon!

Here’s a look at Experimonth: Mood registrations over the four days we recruited participants.

We announced the project on Monday at 10am through updates to Facebook and Twitter, hoping to get 40 non-museum staff registrations by Friday. Right away were invited to do guest posts for Quantified Self and Scientific American (thanks to Nina, Paul, Bora and Anton for their parts in that).

We had a couple days to prepare for the potential influx of attention, so we made calls to and tested our service providers and readied the site for a pay option of $6.47 to participate (we didn’t budget for more than 200 users and that’s the cost per).

Here’s how the community grew one registration at a time over the four days:

Day One: We announced the project at 10am and had a bunch of new users right away.  Registration was steady throughout the day, Facebook referrals (204) were three times that of Twitter (65).  Our highest day by one registration, we added 53 participants. We had 17 participants (beta testers and staff) prior to announcing the project.

Day Two: We didn’t do anything to promote the project on day two in preparation for the potential number of signups through QS and SciAm referrals.  Registrations were slow in the middle of the night and slow, but steady throughout the day.  We added 19 new participants. 

Day Three: We began the day by load testing our server to make sure we could handle a lot of traffic if it came our way. This testing inadvertently broke our registration form and when I updated Twitter and Facebook for locals to signup before the 200 free registrations were filled, the site was down and we didn’t know it.  You can see the window of dotlessness on the chart. We fixed the form and told folks of our error, but I’m sure some registrations were lost.  

Around 3pm, both the QS and SciAm posts went live, but traffic did not balloon like we anticipated.  We ended up with 53 referrals (visits, not registration) from QS and 36 from SciAm over two days.  We added 42 more participants on day three.

Day Four: We updated Twitter and Facebook with “there are still free spots” messaging and got a pretty good bump.  I’m unsure if there’s any significance to the gap between 2-4, but the traffic of the site was a little low then, too.  We added 52 participants.

—

At the end of the four days, we ended up with 166 new registrations.  Our total number of participants is 183, 160 of which are not museum staff (four times my original goal). The site’s traffic analytics show 703 unique visits in those four days, so nearly a quarter of the visitors actually signed up.

The press was exciting, but didn’t really wallop us like we feared/hoped.  We certainly have more participants as a result, but the project remains largely local and manageable in size.

###

Here’s the source data if you want to reference/play with it.

I vowed to keep Experimonth: Mood registration as simple as possible by only requiring name, phone number and email to sign-up. 

The name because I have a theory that communities where folks use their real name have more accountability.  In this way, all participants go by their first name and last initial except for co-hosts (full name) and readers who are not participating (whatever they input on the comment form).

The mobile number and email were required for the text messaging and weekly mission communications.

The downside is that I don’t have too much information about the participants.  I hope to fill in the gaps with questions on the blog and requests to fill out profile information, but the only thing I’m assured is name, phone, email.

This graphic takes advantage of area codes — an implicit but likely inaccurate location measure given the numbers are mobile.  It’s still interesting, and I hope it will prompt the group to put their zip code info in their profiles so I can more accurately chart their participation.