What we learned from the ERES 2023 Conference

Blended learning is going through a transition period

There was limited usage of online learning approaches pre-pandemic. During the pandemic however, everything changed! Everyone had to move fully online in a flash, and scramble to put together resources for their students. As things transitioned back towards normality, perceptions of online learning were mixed, with some students wanting to learn online, and others wanting to be on campus.

As we get to 2023, everything’s changed! Students say they want online learning, but engagement with online learning drops off a cliff when students are presented with the same resources that they engaged well with during the pandemic.

There are however, several key learnings from the blended learning panel that are valuable to share that indicate that blended learning is not only here to stay, but it's going to provide increasingly valuable learning experiences to students:

‍

1. Recorded lectures aren’t engaging students
2. Shifting more passive learning online, and using classroom time for more active learning activities such as discussions and working through problems as a class works very well, and proves to be highly engaging.
3. Online group work is effective, and students remain engaged, but they need to be taught how to work effectively in groups
4. Students need to be very clearly signposted to what to spend time on. Online active learning exercises and quizzes that are graded work well here, as does clearly informing students that content from online learning resources will be covered in exams.

Perceptions of Generative AI, and its potential impact on education, are changing!

‍4 months ago, we were at the ARES Spring Conference in San Antonio. The audience was asked how they feel about Chat GPT & generative AI tools. The majority saw them as a threat, and resisted their use in the classroom. Now however, the consensus seems to be that they present opportunities for new and better ways to teach and learn, though not without risk requiring the redesign of some elements of instruction and assessment.

Assessment is changing due to the increasing prevalence of generative AI tools.

Online testing is harder to manage. Cheating is easier than ever as a result of AI tools. This means that many assessment methods that have been relied upon for many years are potentially less effective; for one, multiple-choice quizzes are less well suited to a world where ChatGPT can be used by students.

Students have been finding ways to cheat forever. Before ChatGPT, it was essay factories, and before that, students were copying one-another's work. The tools may have changed, but the impact is broadly the same. It may be another 3 years before tools used to detect the use of generative AI (originality index scores) are accurate enough to be useful. In the meantime, assessments will likely have to change, or educators will need to spend more time manually evaluating whether or not a student wrote their own work; can they recall what they wrote? Is the writing style theirs? Is it clunky? Are the references real? And is the data real?

Many educators are going back to holding exams in the physical classroom, with some even reverting back to using paper tests.

At many institutions, authentic assessment is absolutely encouraged. Some educators are taking a step further, and allowing students to use AI tools, within a clear framework. This however, requires a lot of work, and is very hard to manage when class sizes are large!

There's no single solution, but we'll likely see a lot of change to the way in which we assess students in the next few months.

There are some big benefits of AI that are starting to emerge

Working with tools like Blended Teaching has allowed some educators to move faster in the classroom. Chat GPT could potentially be used in a similar way to provide individualized feedback to students in a way that isn’t possible to do as a professor with a large group of students. It can help decrease some of the individual workload for educators but while still being authentic and true to students.

These tools are useful for students for whom English is a second language

As educators, we want to know students’ understanding of the concepts, and how well they are able to apply them. But sometimes it can be hard for them to translate that knowledge into a form that's as coherent as what would be created by a native English speaker of the same skill level.

This is where Generative AI can function as a powerful tool for levelling the playing field.

Interesting new teaching methods are emerging too!

A great example is making ChatGPT a student in the class, which appears to be a powerful tool for engaging students. An example that was shared was to have ChatGPT complete an assignment, and then create a discussion in the classroom where students evaluate and critique its work; identifying what they’ve done well, and where there are errors etc.

We need to be careful of biases!

These tools are built on data that’s hidden in a black box. There are countless biases that will exist within these datasets that are entirely unknown. Teaching students how to manage these biases is critically important!

There’s an important issue related to equity to consider

Wealthier students will likely have access to higher quality, or more powerful, AI compared to other students. Better access to tools shouldn’t influence one student’s grades over another's. There’s not a clear answer for how to approach this challenge yet, and it will be an important point to address sooner rather than later!

‍

Many thanks to all of the contributors and panelists. Particular thanks go to Eamonn D'Arcy, Fernanda Antunes, Saul Nurick, and Steve Hood.