Week 1
- Created plan for conducting research
- Looked for relevant literature
- Created template for fake computer science course page
- Attended HCI lab meetings
Week 2
- Worked on and submitted IRB proposal
- Created study interfaces, one stereotypical and one non-stereotypical
- Started looking into recruiting participants
- Booked an experiment space
- Analyzed data for a separate, related project on motherhood bias
Week 3
- Lots of new interns starting this week
- Started working on Qualtrics survey
- Created recruitment materials (posters)
Week 4
- Contacted participant populations (looking at Stanford summer programs for highschoolers)
- Danaë did a power analysis to figure out how many participants we need
- Still waiting on IRB
- Contacted Sapna Cheryan for original study materials on ambient belonging
- Piloted Qualtrics survey around the lab
Week 5
- Pre-tested our websites on Amazon Mechanical Turk to ensure that interfaces are indeed stereotypical and non-stereotypical
- Got original study materials from Sapna, incorporating them into our survey
- Wrote pre-registration for Open Science Framework
- Received IRB feedback to update our IRB proposal
Week 6
- IRB proposal finally approved! We can start sending out recruitment materials and running the study
- Piloted the survey on MTurk
- Piloted the full in-person procedure
- Figured out method of compensation for participants
- Wrote analysis scripts in python
- Booked rooms for running the study
Week 7
- Began running participants in lab
- Because we aren't sure we'll receive enough participants, we decided to also run the study online
- Launched Facebook ad to recruit 16-20 year olds to participate in the study, currenting analyzing this data
Week 8
- Data from Facebook participants turned out to be low-quality; we got a lot of respondents but there were a lot of people trying to cheat and take it multiple times
- Facebook participants were 70% female, male populations did not randomize well
- Facebook data shows significant differences in DV's by gender but not condition, but this data isn't really valid anyway
- Figuring out next steps for getting better data, will probably run on MTurk on 18-25 year olds
- Still running in-person participants
- Started writing a CHI paper draft
Week 9
- MTurk data yielded positive results that supported our hypothesis!
- Saw differences by gender for most DV's and also by condition
- Thinking about how to extend this work and looking at the big picture
- Planning on building a quantitative model for inclusive design to validate on real life interfaces
- Working on CHI short paper
Week 10
- Finished up first draft of CHI paper
- This was my last week, and I'm glad to say I had a really great time this summer at Stanford. Thank you Danaë, James Landay, and CRA-W for providing me with this opportunity!