Who's teaching AI?
Back in March, I shared the results of a survey of middle and high school students about how they use AI in school. The bottom line: They pretty much all use AI, most of them are not using AI to plagiarize outright, and middle schoolers push the envelope with AI further than high school students.
But what about their teachers?
If Abbott Elementary can dedicate an entire episode to this, it's definitely worth a nerd processor case study!
Survey says: Bring on the AI
My student survey focused on people based in the greater Seattle area, and the teacher survey does too. In order to keep symmetry with the student survey, this case study includes responses from 100 middle school teachers and 100 high school teachers. They teach math, science, English, history, or foreign language.
Some of the same caveats from the student survey apply here. Seattle is a tech-forward market, so it's reasonable to assume that participants are on the farther edge of the AI adoption curve. That being said, while AI adoption in other regions may be lower, we can still learn something from the directional trends in this data.
tl;dr Teachers are using AI. A lot. 80% of respondents have used AI in their teacher jobs.
School AI policies: Pretty centrist
In the student survey, middle school students showed more progressive attitudes about AI than the high schoolers. Middle schoolers were more likely to disagree with their school AI policies, more likely to have used AI without their teachers knowing, and less likely to consider it cheating to use AI. Middle school and high school teachers showed no such differences.
Just over half of the teachers agree with their school's AI policies; the rest are split between thinking the policies are too permissive and too restrictive. In other words, schools have found a policy spot that is pretty centrist as far as teacher opinion goes.
I was interested to see that almost 70% of teachers encourage their students to use AI in at least some circumstances. It's worth calling out again that Seattle is a tech-forward market and I'd expect national (and international) numbers to be lower, but I found this really striking.
Teachers are not just specifying whether students can use AI on a particular assignment; 60% of them are providing structured guidance on where, when, and how AI should be used. I find this data point extremely encouraging. This shows remarkable ability on behalf of educators to adapt to a rapidly evolving technological landscape, and willingness to help students build the skills they will need down the road.
Teachers use AI. A lot.
The guidance that teachers provide their students about how and when to use AI may come from personal experience. 80% of the teachers have used AI in their teacher jobs at least a little, compared with 81% of the middle and high school students. I'd expected teacher adoption rates to be lower than student rates, but they look the same.
As for where teachers are using AI in their work, nearly half the teachers are using AI to get ideas for lesson plans, and almost a third are using AI to help write assignments.
I was surprised and a little disappointed to see that so many of the teachers are using AI to write student feedback. Granted, I am deeper in this topic than most, given that we built an entire AI product at Textio focused on writing fair workplace feedback. But the bias in using off-the-shelf AI like ChatGPT to write feedback is pervasive, and this use case poses real risks for equity in schools.
Our moment in the metaverse
I'm not sure that anything summarizes the current inflection point better than teachers using AI to check whether their students are using AI. This was the most common example that came up in the survey when teachers explained the types of plagiarism that they're using AI to uncover.
Using AI to spot AI is about the most meta thing ever, and that's our current moment.
What do you think?
Thanks for reading!
Kieran
p.s. I recently came across Emily Parkhurst's Formidable newsletter, and specifically this story about how women don't choose lower-wage jobs; wages drop when women enter an industry. This is exactly my kind of data story, so if you like nerd processor, you might like this too!
Catch up on nerd processor case studies | Subscribe | nerdprocessor.com