Using Web analytics to improve education results
Hundreds of thousands of students have already tried our Hour of Code tutorials and the follow-up online course. Using online instruction with “self-grading” puzzles, we get the benefit of measuring student performance on a puzzle-by-puzzle basis.
This graph shows “completion rate” and “dropout rate” across every puzzle in our curriculum. In other words, the percent of students who tried and solved puzzles, and those who tried puzzles, but dropped out and never returned to our system.
Puzzles with low completion rates or high dropout rates are the problematic ones we still need to fix.
Note: this data isn’t brand new. We’ve already fixed bugs to address the most glaring issues you can spot on the graph.
How cool is that? We can use Web analytics to identify when our tutorials aren’t helping students. Soon, we can run A/B tests to improve childhood education results.
Coding isn't only for wizards and super-geniuses...
Check out what Code.org is up to!







