If you plan to self-teach these materials, it’s important to understand what each piece is and how to best use it.
There are four pillars to successful data science self-study, resources for all of which are provided on this website:
- Learn basic coding and practice on an interactive online coding platform that provides real-time feedback.
- Translate key concepts learned on the coding platform to the social science context.
- Complete an applied project on your own using the skills you learned.
- Watch a video explaining how to complete the project and compare what the “expert” did and what you did.
That means you should organize your learning by topic (in a formal course, this would be a week of material) and do four things for each topic:
- Complete the DataCamp courses and do the readings associated with that topic.
- Watch the lecture video.
- Download the project assignment PDF and complete it on your own.
- After you’ve finished, watch the Debriefing video. Compare what I do and what you did. There are often multiple correct approaches to any given problem, and thinking through why you did things differently from me is a critical step in learning to program.
Remember to pace yourself. The modules here were designed as weekly content for a graduate level course, i.e., at least 10 hours a week, upwards of 15. Think of this as roughly the same investment of time and effort as learning graduate-level statistics the first time.
You are literally learning a new language, one which will allow you to speak to a computer. It will take some time. There is vocab, syntax, and grammar to learn, and you will only become fluent if you devote a meaningful amount of time to it.
If you can only devote 2-3 hours a week, schedule yourself to complete one module a month. It’ll take a year, but then you’ll be an intermediate R user, which is significantly more advanced than most social scientists!
If this sounds good to you, browse the schedule and materials to get started.