What we learned while teaching Python and Data Science

Pedagogy and lessons learned from teaching introductory Python and Data Science courses online.
This is how we approached the matter, what we learned and where we want to go next.

Transcript

1. WHAT WE LEARNED WHILE TEACHING PYTHON AND DATA SCIENCE Saul Diez-Guerra O P E N D A T A S C I E N C E C O N F E R E N C E_ BOSTON 2015 @opendatasci
2. WHO ARE YOU?! Saul Diez-Guerra Spaniard in NYC EE/CS background CTO @ Thinkful I (might) have targeted you in the past Payback! You can target me: saul@diezguerra.com @diezguerra @definitely 2
3. THANKFUL? 3
4. Thinkful is a mentor-led, project-driven, self-paced online tech school that gets (aspiring) software professionals job-ready. 4
6. Thinkful is a mentor-led, project-driven, self-paced online tech school that gets (aspiring) software professionals job-ready. 6
7. N_GRADS > N_JOBS 7
9. CREATIVE/COGNITIVE > * 9
11. MILLENNIALS 11
12. MILLENNIALS! Ö 12
15. EXPECTATIONS ARE SHIFTING 15
17. PREACHING, NOT PITCHING We're funded, but listen, ye guild. Resources, conferences like this will help. EdTech is booming: break through? 17
18. SO FAR… 4000+ students ~20% Python & Python for Data Science 18
19. ORIGINAL APPROACH MVC: Recipes and formulas. Missing baseline. Grok the use cases, grok the tools. 19
20. DS IS A BROAD TOPIC Wide background of students. Different abilities that need attention. 20
22. THE POWER OF THE MENTOR 22
23. All about that mentor-led experience. Gauge where an inbound student is: steer, recommend and set expectations about progress, industry, career and job market. Community stewards. 23 THE POWER OF THE MENTOR
25. CURRICULUM The case of the experts. Building content: cost and alternatives. Lack of prior exposure to programming has a solution. Drop-out “by frustration” can be avoided. Also, git. Lean feedback. MacCready! 25
26. CURRICULUM Pre-testing: pop quiz! http://www.nytimes.com/2014/09/07/magazine/why-flunking- exams-is-actually-a-good-thing.html 26
27. CURRICULUM Project-based. Here’s a tool. Here’s a problem. Your plane to fly. Not a problem set, but a challenge. 27
28. CURRICULUM 28 vs vs
29. CURRICULUM Citibike project. Pirate bartender project. 29
30. CAPSTONES! https://github.com/amaner/waveforms https://github.com/wiharto/Thinkful---Capstone/blob/master/ Digits%20Recognition.ipynb 30
31. NEXT(CURRICULUM) • Some improvements • ML fitting • Dataviz • Communication • More face-to-face • Intensive: the path from core • Office hours: experimental dynamics • Ultimately, reducing MTTM, less frustration, less overwhelmed students 31
32. NEXT(CURRICULUM) • Modularity / Paths • Graduates and seniors need less hand-holding and encouragement, as they fear nothing. • Fast access to discrete topics. 32
33. RECAP Macroeconomic job trend. More education: flexible, accessible, project-based. Data Science is too broad, but mentors help a lot. As specialization is demanded, we modularize while conserving the original precepts. 33
34. WHAT WE LEARNED WHILE TEACHING PYTHON AND DATA SCIENCE Saul Diez-Guerra T H A N K S !