Let's get meta! What the data from the joind.in API can teach us about the PHP
summary
Joind.in data get scraped into a MySQL DB using Node 10 with TypeScript and Sequelize, yielding ~2,500 events and gender guesses for 95% of users via Genderize/Gender-API; analysis finds a ~5:1 male:female speaker ratio, women average 13 comments vs 6 for men, and a logistic regression shows male commenters are ~0.25 as likely to give a high rating to female speakers while female commenters are ~3× more likely to rate female speakers highly — results drive concrete UI and rating-scale recommendations.
AI-written summary · about
chapters (13)
- 0:00Start
- 4:09Where's the Code
- 6:16Process for Scraping and Cleaning the Data
- 9:38Data Challenges
- 11:52Guessing Gender Based on Name
- 14:56Data of Results
- 16:02How Many Comments Do Women Female Speakers Get on Their Talks Compared to Men
- 16:44Word Clouds
- 19:52Gender Bias
- 20:50Controls
- 24:05Constructive Feedback
- 24:56Ui Tweaks
- 25:14Changing the Rating Scale
related talks
49mHacking In-Group Bias for Fun and Profit
31mFrom Bias to Belonging: Creating inclusive Tech Communities
15mBechdel.io: How We Used JavaScript To Help Make Film More Inclusive
32mBuilding inclusive Open Source communities
39mState of diversity in open source communities
53mCoding Like a Girl
55mDiversity in OpenSource … show me the data!
37mExtracting Insight From A Muslim Marriage App