I had to skip a progress update last month, as I had health issues.
- Outreachy accepted 49 interns in the December 2019 round
- The experiment of separating the initial application period and the contribution period was a success!
- Wrapped up the first Outreachy longitudinal study to collect statistics on Outreachy alums
- Wrote a blog post categorizing Outreachy projects by topic
- Updated internship documentation
- Updated organizer documentation
- Improvements to the Outreachy Django administrative back-end and view the organizers use to track interns
December 2019 round
The application period for the December 2019 is closed. We announced we accepted 50 interns this round. One intern was later found to be ineligible for Outreachy. Still, 49 interns is our biggest round yet!
In the December 2019 round:
- 1,774 applicants filled out an initial application
- 590 initial applications did not align with Outreachy’s goal of supporting diversity in free software
- 532 applicants were ineligible for Outreachy due to full-time commitments during the internship
- 91 applicants were ineligible for Outreachy due to not being able to be paid by a U.S. non-profit
- 607 applications were accepted into the contribution period
- 236 applicants recorded starting a contribution
- 206 applicants created a final application
That number of people who recorded a contribution this round went up! Most rounds, we have around 200 people who record a contribution and about 175 who complete a final application.
That means that applicants who completed an initial application and then had to wait to start the contribution period didn’t lose interest. That means that separating out the eligibility checking and contribution period was a success.
One Outreachy community was caught off guard with the new application process. They didn’t know they needed to get their project description in early and promote their project before the initial application deadline. That community ended up not accepting any interns this round. The data could be coincidence, because there have been multiple rounds in the past where that community didn’t accept an intern.
There were a few communities who normally participate that didn’t this round. That was more than made up for by the new communities who joined. We typically have around 65 projects each round, and this round we had 83 projects listed.
All in all, a success and a growth in the number of accepted interns over last round.
In October, we launched our first longitudinal survey. The goal was to find out statistics about how successful Outreachy was at encouraging our past interns to contribute to free software and stay in the technology industry. 230 out of 589 alums responded to the survey.
To incentivize Outreachy alums to fill out the survey, we offered a reward of an Outreachy sticker if they completed it. 169 of the alums wanted to receive an Outreachy sticker for filling out the survey. The rest of the respondents didn’t fill out the address field and didn’t want a sticker.
Part of the reason for running the longitudinal survey was to understand how much time and effort was required to send a small reward (an Outreachy sticker). This would let us know where the pain points were for a future Outreachy fundraiser with physical rewards.
Stuffing envelopes was a breeze. It took four volunteers about 1.5 hours to stuff 169 envelopes. (We had a wet sponge on hand to seal envelopes to save volunteers’ tongues.)
The stuffing was fast because we chose to not hand-write the thank-you cards. We printed the thank you message on 2″ x 3″ stickers through Moo.com, and put them on the inside of 4″ x 6″ thank you cards. The 1″ x 5″ Outreachy stickers (also printed through Moo.com) fit fine inside the thank-you envelope. Printing stickers to put on the inside of thank you cards was more cost effective than printing messages on thank you cards through other printing services.
The real time sync comes in addressing and mailing the envelopes. The (untested) method I settled on was printing pre-paid stamps through stamps.com. It requires a monthly subscription, and some of the features like taking addresses from a spreadsheet may only be possible with the use of Windows software. We’ll see how it goes.
The time-consuming part comes in normalizing the addresses from the longitudinal survey into standard international format for each country. I quickly skimmed and recorded each address’ country, then sorted the spreadsheet by country. Normalizing 56 international addresses took about 2 hours.
Finding a resource for standard country address formats was hard, but this one seems good. It’s untested whether it’s correct.