Ethical Tech YVR

AboutResourcesSessions

Blog Posts

No results for undefined
Powered by Algolia

Learning Analytics - Futures of Learning

Photo by Element5 Digital on Unsplash

This month I had the privilege of presenting at Stella Lee's Futures of Learning meetup group alongside Alec Balasescu. The theme for the session was "Learning Analytics". As an instructor at a web development bootcamp, I work closely with the education team in gathering and analyzing performance data to improve our curriculum and the general student experience. Here are the rough notes that I used for my presentation. Soon I'll compile these notes into a nicely formatted blog post!

Topic: Ethics of Learning Analytics - Student Data & Performance Prediction

Details:

Learning analytics has the potential to enable educators and learning designers to gain insights of their learner's needs and to use that understanding to improve the learning experience and provide intervention. Yet collection of student data and to use said data to predict their performance and influence their learning behaviors come with a number of ethical challenges, include learner privacy, interpretation of data, assumptions and bias in the type of data one collects, anonymization of data, and informed consent.

Presentation Notes

  • What does learning analytics mean to me?

    • Collecting data on assignment submissions, student-teacher interactions, student feedback, and student behaviour (e.g. showing up on time, how long it takes to submit assignments)
    • This data could be used to:
    • decide How to distribute resources (giving extra/less support)
    • Predictors for difficulties that particular students may face
    • Aid in decision-making for offering scholarships, and spots in programs
    • Use of data to assess teacher performance
  • What are some of the positives?

    • Identify hotspots of issues in curriculum
    • Keep an eye on student performance
    • At-a-glance information available
    • Possibilities for personalized curriculum
  • What are some of the issues?

    • Increased stress on students
    • Labelling a student early on as "needing support" can be a self-fulfilling prophecy, if the student comes to feel that they require that extra support (learned helplessness)
    • Fairness of resource distribution'
  • Things to keep in mind

    • Training people to make fair assessments (what should the bar be?)
    • Ensuring that assessment is FORMATIVE as opposed to SUMMATIVE
    • To make valid conclusions, we need a LOT of data! Be careful not to generalize off of a small dataset
    • Constantly check your assumptions
    • Do I feel this way about a student because they exhibit some "recognizable patterns", and am I attempting to use data to validate that?
    • How much should you disclose to students about the information that’s collected, and how it will be used?
  • Highlight some resources for further reading, and bring examples from literature

    • Weapons of Math Destruction

Quote: "As education becomes increasingly scaled and asynchronous, analytics becomes more important as a tool in support of high-quality teaching and learning practices that are responsive to meet the needs of students in a timely manner" - Timothy Harfield, Senior Product Marketing Manager for Blackboard Analytics (https://elearnmagazine.com/learning-analytics-possibilities/)

  • Pose questions/challenges to be discussed in smaller groups
  • Who should be granted access to seeing student performance data?

    1. e.g. which members of an education team? the students themselves?
  • How much should you disclose to students about the information that’s collected, and how it will be used?
  • Where should data be stored?

    1. Needs to be in the country that the data was collected
    2. Who is the data on? E.g. children...
  • If you're collecting to a school MOOC, you need to look at what your security for your app that's hooking into that data (e.g. if you're connecting to it remotely)
  • Data being provided externally, to someone who doesn't have the ability to see the context, these people may make incorrect or dangerous extrapolations... we need the right people to do the analysis

Thoughts

Presenting at the meetup was a fantastic experience, and I hope to get to do something like that again soon. I know the notes above are a bit incoherent, but I'll compile them into something legible in a while.

I wish I had made more time to put an emphasis on how I think learning analytics can put us at risk of putting students into "buckets" or profiles, and the issues that arise out of that... but I still need to sort those thoughts out a little more before I'm comfortable putting it out there.

Cheers!