With the support of the William and Flora Hewlett Foundation, Open Education Solutions announced the winners of the $100,000 Automated Student Assessment Prize (ASAP) this morning in Washington DC. The top three teams, made up of international data scientists, developed predictive algorithms that were able to score thousands of essays very quickly with remarkable accuracy compared to expert graders.
Five important conclusions can be drawn from the first phase of ASAP:
- State aspirations for affordable tests that incorporate more writing and faster feedback are well founded. The two consortia developing new online tests will be able to incorporate a significant amount of writing to better measure deeper learning and progress toward college and career readiness.
- Tech tools will leverage teacher talent. Teachers can expect more writing from students when they have powerful tools that provide instant multi-trait feedback.
- Talent is global. The competition attracted data scientists from around the world and, as the winning team illustrated, many self-organized into super efficient teams.
- Prizes efficiently accelerate and focus innovation. The Hewlett Foundation funded prize mobilized talent and investment achieving a significant advance in the sector for a relatively modest investment.
- Platforms scaffold innovation. The Kaggle.com platform and new data mining tools make it possible to attack mountains of data and extract new understandings and powerful capabilities.
This competition follows a February vendor demonstration designed to gauge the state of the field. The study can be found at http://bit.ly/HJWwdP
Next in the ASAP sequence is short answer questions—a more difficult scoring challenge.
The full press release and a description of the winning teams can be found here.
For more, see these GettingSmart posts: