I am starting a thread on education responses to chatGPT. Please feel free to add to it.
I am not really concerned about the reaction posts, I am looking for useful information and the impact.
@martin
https://www.openedtech.global/blog/open-edtech-news-2/thoughts-about-the-impact-of-ai-text-on-assessment-9
John Spencer https://spencerauthor.com/englishai/
Steve Heimler
https://www.youtube.com/watch?v=jvNm7aXYPJ8&ab_channel=Heimler%27sHistoryTeachers
@scerruti @martin @edutooters I was talking about this with my wife last night. I think the potential for cheating is more of a symptom. Students feel the need to cheat because we've set up an education system that prioritized grades and rankings over actual learning.
The real solution is to revamp how we evaluate students, but that requires a big shift from secondary -> college.
@scerruti @martin @edutooters If students felt like school was about *actually learning* instead of competing for GPAs and limited slots at expensive colleges (and therefore scholarships), maybe they wouldn't stress out when it takes a little longer to grasp a concept.
It doesn't help that we need to set quarterly or semesterly deadlines for completing work.
Also important to note that in *most* lines of work, you're never graded. You just decide when the product is good enough and move on.
@aiannazzone @scerruti @martin @edutooters also important to note that in many of not all work situations one works as part of a team, it doesn't matter if it's in a restaurant, IT, teaching, HR, retail, etc... Individual accountability is often hard or impossible to ascribe because of complex contexts and interdependence on others.
@ZingerLearns @aiannazzone @martin @edutooters "Individual accountability is often hard or impossible" as someone who has been a manger for more years than a teacher working with complex multi-site teams, I can tell you this is not impossible. It's made difficult in education by the lack of authentic assessments and the underlying design of modern education.
@scerruti @ZingerLearns @martin @edutooters I wish we could capture & assess the natural progression of coding fluency. Similar to learning a new spoken language, you start out stilted, but slowly your elegance and nuance improves.
What may have once been a complicated mess of nested if/else-if statements may later become a few methods with careful use of return statements.
But that's just so hard to quantify on a 0-100 scale (and in a way that adjusts expectations as the students progress).
@aiannazzone @scerruti @martin @edutooters I think the nuanced nature of individual progress is an important point here. Not everyone develops a particular skill or disposition the same way at the same rate.
I think there is an important question on what counts as sufficient or adequate or acceptable progress and learning?
I also think about examples in programming where some measure success as lines of code in a given period of time, while largely or entirely ignoring the quality or efficiency of the code.
All of these are great reasons to move to #ungrading
@ZingerLearns @scerruti @martin @edutooters And to get back to the original question: grading/evaluating based on those kind of metrics would help buffer against tools like #ChatGPT. I don't care if your code is perfect right now. I just want to see meaningful progress.