There is already a video introducing Bayes’ Theorem in the probability section, and a similar example is addressed in another video. This could be expanded greatly; possible topics include the odds formulation with likelihood ratios, updating, a discussion of priors, and Bayesian inference. (The last one might be better off as a series in the statistics area instead.)
We can also aim for more than just video lectures, although that would be a very good start. Khan is developing adaptive software which randomly generates exercises and tracks the student’s progress with the concepts those exercises test. Creating such exercises for Bayes theorem would help people develop an intuitive understanding of how to use it, far beyond what they could get out of just watching the videos.
Creating good exercises for traditional concepts that form the building blocks of rationality is probably the best way to go about this. Once you’ve shown that you can create engaging exercises, and have formed relationships with people at Khan Academy, your suggestions will have more weight and you can create more experimental type exercises. Sal can only create about 10 videos a day, and the requests he gets from teachers using Khan Academy in actual classrooms (or from the people giving him funding) probably trump those he gets from people on a rationality blog.
For anyone reading this who wants to create an exercise teaching Bayes’ Theorem, first read the Github Wiki. You can also get help from Khan Academy employees and volunteers in the HipChat room.
I’m an ex-intern at Khan Academy (who worked quite a bit on the exercise framework) so feel free to email me any questions or to get me to look at your pull request. My email is jp@julianpulgarin.com.
A specific proposal:
There is already a video introducing Bayes’ Theorem in the probability section, and a similar example is addressed in another video. This could be expanded greatly; possible topics include the odds formulation with likelihood ratios, updating, a discussion of priors, and Bayesian inference. (The last one might be better off as a series in the statistics area instead.)
We can also aim for more than just video lectures, although that would be a very good start. Khan is developing adaptive software which randomly generates exercises and tracks the student’s progress with the concepts those exercises test. Creating such exercises for Bayes theorem would help people develop an intuitive understanding of how to use it, far beyond what they could get out of just watching the videos.
Creating good exercises for traditional concepts that form the building blocks of rationality is probably the best way to go about this. Once you’ve shown that you can create engaging exercises, and have formed relationships with people at Khan Academy, your suggestions will have more weight and you can create more experimental type exercises. Sal can only create about 10 videos a day, and the requests he gets from teachers using Khan Academy in actual classrooms (or from the people giving him funding) probably trump those he gets from people on a rationality blog.
For anyone reading this who wants to create an exercise teaching Bayes’ Theorem, first read the Github Wiki. You can also get help from Khan Academy employees and volunteers in the HipChat room.
I’m an ex-intern at Khan Academy (who worked quite a bit on the exercise framework) so feel free to email me any questions or to get me to look at your pull request. My email is jp@julianpulgarin.com.