Monday, November 24, 2014 KADataScience 4 Comments

Welcome to the Khan Academy Data Science Blog!

The mission of Khan Academy is to provide a free, world-class education for anyone, anywhere. On the Data Science team, we’re particularly focused on understanding the “world-class” aspect of the mission. For example, how should we define “world-class”? How do we measure quality? And if we can measure quality, then how can we systematically improve it?

In general these are very complex and messy problems. In fact, it’s impossible to boil a world-class education down to a simple formula or algorithm. But at the same time, online learning does afford us new opportunities to study how people learn (and teach!) effectively. If we can make just a little progress on understanding which videos teach a particular concept most effectively, or how to keep students engaged and motivated while working on tough material, then we can share that knowledge with the world and start steadily piecing together a fuller picture of a world-class education. It’s just like science -- data science!

In order to make progress on our key questions, here are just a few of the things we typically work on that you’ll be able to read about on this blog.

    • Developing metrics for measuring “learning gain” to estimate how much users are able to learn while using Khan Academy.
    • Designing and running randomized controlled experiments (often called A/B tests) to improve our:
      • Video, exercise, and article content
      • Personalized, adaptive learning algorithms
      • User interface and overall user experience
    • Analyzing data alongside selected partners and educators to understand and improve efficacy of Khan Academy usage in schools and classrooms.
    • Building an adaptive assessment engine and content, and validating those assessments with respect to widely accepted standards.

We hope that sharing our successes (and our non-successes!) on all of these fronts will provide a useful and interesting source of information for educators, researchers, and the simply curious. Khan Academy has certainly benefitted from the collective wisdom found in academic literature, technical blogs, and feedback from our user community; we now look forward to sharing back details of our own findings with those communities.

Enjoy the blog!


  1. Hooray for Data Science! I'm looking forward to the Khan Academy perspective on a topic near and dear to me.

  2. As Khan academy freely shares so shall I share. When we phrase our mission statements in the "current affirmative" we realize the actualization of the mission now; i.e. "our mission is achieving" rather than "our mission is to achieve" which puts the actualization of the mission in the perpetual future.

  3. We do not own the future, nor the past, only the here and now, so ... but we can use math to get an idea of the future, by learning from the past. Those who's heads are always in the past, or the future rarely have joy. "Achieving" implies the perpetual present.

  4. Really Cool! Does this mean KA will be doing PowerPivot and Hadoop videos? ;)