Analytics with Magnus

Cortexio has had a research focus since the start, with the intent on improving the student efficiency by leveraging data gathered from the study sessions.

Magnus Ödman has been working full time with this since 2018 at Cortexio. Here are some thoughts from Magnus.

(Cortexio is the learning platform created by Hypocampus AB.)

Background

Education:
Master in Engineering Physics at Chalmers University.
Career:
IT consultant for over 15 years with a wide range of assignments, most recently at Volvo Cars.
On his spare time:
Plays guitar and sings in a band, plays old boys football and cuddles with his dog.

Work on Cortexio

Magnus has developed a new Analytics service for Cortexio that collects and processes study activity. This service makes it possible to give better recommendations for each student on what to study and which questions to repeat.

How much data do you gather on a normal day?

We log 500 000 student interactions every day. An interaction can be a student reading a chapter in a book or answering a question. The interaction analytic data has doubled since last year.

What are you working on right now?

We are refining the  spaced repetition algorithm for a better user experience. I think we nailed it for our new users but we have to improve the experience for long term users.

What does a normal day at the office look like?

In the morning we have a standup where we discuss what we plan to achieve that day. Then there is a lot of thinking, discussing, programming and coffee drinking.

What will you work on next?

Cortexio is expanding with new partners. In the next 6 months we want to offer them analytics services to help them build successful sites giving their users the best possible learning experience.

Study time 2019

Focused study time
Focused study time

2019 we set a record of the total number of study hours in Cortexio! 182 thousand hours were spent studying, that’s a 48% increase compared to 2018.

A student spends an average of 5min 30s on each chapter and 1min 20s on each question. This is what we call focused study time – which is about 15-20% lower than the total time in the system – since we only measure the time a student is actively studying.

Having good precision in the measurement of the time spend studying is important for the research conducted at Cortexio.
We collaborate with researchers at Linnaeus University and Umeå University to investigate what successful study patterns in Cortexio look like. We collect a big number of data points during a study session: Reading time, time spent on images and videos, text markings and notes made, time spent answering a question and reviewing the answer, etc. In addition, the researchers have access to a cognitive test that a large number of students voluntarily answered, as well as diagnostic knowledge tests.

Hopefully, the research can form the basis for a more adaptive study experience.

For example, the data we collect is used to:

  • Adjust the relationship between reading time and question time based on the individual’s cognitive profile (currently the distribution between reading time and questions is 50/50)
  • Individualize repetition patterns
  • Customize the difficulty of the questions to the user’s level of knowledge

In the spring of 2020 we will work on improving our interface to repeat questions, please come up with suggestions on how you would like it!

Feedback in Cortexio

Each month we receive hundreds of feedback messages concerning improvements on the content in Cortexio. We have a living material and change it according to new best practices and guidelines in the different educations.

On a single month we receive around 400 feedback messages through the Feedback function. The feedback is always read, and we and the different sites on Cortexio try to reply to as much as possible. Thanks everyone for giving feedback!

:blush: