Learning Engine Improvements

Written by Neesha Desai

At Shoelace, delivering high-quality, educational content is a top priority. Every change to the game is balanced against how it will impact the players’ learning experience. One of the biggest adjustments to the platform we’ve made over the past year, was the redesign of the question window itself, aiming to make the reading of passages and question answering experience smoother. 

However, some of the changes we’ve been making dive much deeper than those visible to educators and players. For example, we’ve recently completed a Diversity, Equity and Inclusion audit of our passages and questions and, as a result, we’ve updated our content to reflect our commitment to be inclusive and culturally responsive. You can read more about our update from 2021 here (an updated report for 2022 will be released before the end of the year). The other big area we’ve been working on is our learning engine.

Updating The Learning Engine

The learning engine is responsible for choosing what questions and passages a player will be given next, during regular game play. While over the past few years we’ve tweaked it in small ways, we’ve also been toiling away at a much BIGGER update. Because this update is so big, it is being released in phases. Phase one has been in testing over the past few weeks (thanks to all of those who’ve helped with that), and we’re super excited to announce that we’ve now enabled phase one of the learning engine for everyone. 

How This Benefits You

1. We’ve changed from reporting on a player reading level to a player’s reading comprehension level. The reading comprehension level (RCL) takes into account not just the readability of a passage (which is mainly impacted by vocabulary and sentence complexity) but also the reading comprehension questions associated with the passage. 

2. There are upgrades to how the learning engine chooses which passage to assign next to be more tailored to the individual player’s progression. 

3. We’ve introduced more unpredictability. The past learning engine had players moving predictably between passages and our standalone skill questions. We know that some students would feel more inclined to “guess” on a passage knowing that they would soon encounter the shorter standalone questions. With this update, players will no longer be able to predict in advance what question types they’ll get next, leading to increased engagement.

These improvements are all available to you now with our phase one release. In the next releases, we’ve got a lot more content coming and big changes to the presentation of players skill progress in the educator dashboard. Stay tuned. 

About Neesha Desai

Neesha Desai is the Director of Product at Shoelace. She has loved video games since she was little, starting with Reader Rabbit. Her passion for technology and games led her to get a PhD in Computing Science. From there, she was the lead architect of one of the top early reading programs (Squiggle Park). Finding new and creative ways to use the best parts of video games to motivate kids to reach their full potential is her happy place.


comments for this post are closed