Feeling the Churn

It’s movie awards season. For me and my family, this usually means heading to the theaters to catch a nominated film or renting ones that we may have missed in the theaters. After Argentina,1985 won the Golden Globe for Best Picture – Non-English Language, I began searching to see if I could find it someplace in one of the streaming services to which we subscribe. I found the film on Amazon Prime Video and gathered my wife and daughter last Sunday night watch to the film.

I don’t use Amazon Prime Video often. We’ll rent a movie occasionally through the service, but I spend a lot more time using one of our other streaming services. Despite this limited use, I navigated the Prime Video menu and selected the film. For some reason, Amazon (inaccurately) predicted that we’d prefer to watch the dubbed version of the movie rather than read subtitles. So, I stopped the movie, fumbled through the captioning menu, and tried to restart the movie with English subtitles rather than the default dubbing. I was somewhat successful. While the movie was now playing in Spanish with English subtitles, the positioning and size of the captioning made it difficult to read. So, I navigated back to the captioning menu but accidentally restarted the movie again. At this point, I started getting frustrated. We had watched the first few minutes of the movie several times at this point. So, when my daughter asked if I needed a hand, I angrily tossed the remote to her. And I sat there silent and angry.

At the time, my daughter wondered why I had gotten so angry. She was graciously offering to help me in a time of need. She streams a lot of movies and almost always uses the captions when she watches. She has much more experience knowing that to access Prime Video’s captioning you have to click a certain button and use the directional remote to navigate to the menu. She also can expertly navigate to the captioning menus in Hulu, HboMax, AppleTV+, and Netflix. She’s a captioning wizard. She is well positioned to offer captioning assistance to anyone in need. So, why had I gotten so angry?

It has taken me a while to realize that my negative emotions really had nothing to do with her abilities or my own. They’re rooted in something else. This morning, I was listening to the ReThinking with Adam Grant podcast and Adam was interviewing Claude Steele. While he has held numerous academic positions at Columbia and Stanford, Steele is best known for his work on stereotype threat. Stereotype threat is defined as a “socially premised psychological threat that arises when one is in a situation or doing something for which a negative stereotype about one’s group applies” (Steele & Aronson, 1995). According to stereotype threat, members of a marginalized group acknowledge that a negative stereotype exists in reference to their group, and they demonstrate apprehension about confirming the negative stereotype by engaging in particular activities. Steele describes this apprehension as “churn.” When someone feels like they’re in a situation where they could possibly be seen or treated in terms of the group, they can feel a mental and physical stress. They’re in a state of churn.

I know it might be tough to read that I’m somehow chalking up my anger to stereotype threat but let me explain. As I’m getting older, I’m becoming acutely aware that others may see me differently. My beard is whiter. My hair is greyer. I’m no longer identifying with the cool, hip, young academics. More often than not, I’m one of the older and more experienced faculty members in meetings. While I blog and podcast and write about technology, I feel I’m increasingly being lumped into the “old guy” category. With that comes my perceptions that others may see me as not being technologically able. That’s a huge identify shift for me. I’ve been the “tech kid” forever. Back in middle school, I was programming on a TRS-80. In high school, I helped to set up the movie projectors in classrooms. I built a microprocessor in college and learned to program in three different languages. Over the last five years, I’ve run technology workshops at my university, at local school districts, and at national conferences.

And now, I can’t use a stupid remote????

Thankfully, the ReThinking episode offered some guidance and some comfort. In their discussion, Adam Grant summarizes the challenges of stereotype threat and offers some ways to navigate the churn.

“The mistake a lot of us make in stereotype threat situations is we think, ‘Okay, if this performance goes poorly, it’s creating a permanent and pervasive signal about my lack of ability. You know that that’s bad. I’m never gonna be good, and I’m never gonna be good at anything.’ And if people learn to make more specific and local attributions and say, ‘Okay, this performance or this test is not diagnostic of my ability. It’s not diagnostic of my ability today. And it’s also not diagnostic of my ability tomorrow. It’s just a, a snapshot of my performance in one particular moment, which happened to be a very stressful, high-anxiety experience,’ it’s a little bit easier then to not dis-identify with the domain, but not overreact to the performance in that moment as representative of the domain.”

So, maybe I can’t always figure out how to turn on the captions. It’s okay.  Thankfully, I have a smart, capable daughter who can help me in a time of need. And that’s nothing to get angry about.

References:

Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual test performance of African Americans. Journal of personality and social psychology, 69(5), 797.

Advertisement

ChatGPT: A Primer

A new semester starts on our campus this week. In preparation for the new semester, my dean, Dr. Lara Willox, reached out and asked if I’d be willing to write a short blurb about ChatGPT that she could include in her semester kick-off newsletter. Since I’ve been writing and reading and talking about ChatGPT a lot over the last six weeks, I was happy to help. She was gracious enough to let me share that blurb with all of you. Enjoy!

Last November, OpenAI released an artificial intelligence chat bot called ChatGPT into the world. While chat bots have existed for years, ChatGPT disrupted the technology and educational worlds. The tool can quickly and effectively generate text-based responses to all sorts of questions. While previous chat bots were subscription-based or only accessible to research communities, ChatGPT was offered for free to the general public upon its release. Within a few days, millions of people around the world were actively using the tool. Since its artificial intelligence based, ChatGPT will continue to learn and improve with more use.

ChatGPT offers some clear challenges for our work as educators. There are already stories nationally about students submitting ChatGPT-generated work as their own. While traditional plagiarism tools like Turnitin won’t flag ChatGPT-generated text, with a little sleuthing, educators may be able to detect whether a student has submitted work that ChatGPT has created. ChatGPT’s writing should be the first clue. Its writing is pretty formulaic; the tool uses similar transition phrases and doesn’t regularly vary its sentence structure. Also, while it is built on a huge database of information, ChatGPT doesn’t always make accurate connections between concepts and will sometimes just make up information or sources.

If an educator comes across a discussion board post or a paper that doesn’t sound like it was written by a student, there are a few tools that can help.
https://openai-openai-detector.hf.space/
https://etedward-gptzero-main-zqgfwb.streamlit.app/

Although neither tool will definitively say whether some text was written by ChatGPT, both will provide metrics that can be used to inform a plagiarism conversation with a student.

While ChatGPT may present some problems in our classrooms, the tool can also offer some opportunities. We can have students analyze the text that ChatGPT generates and have them critique the connections the tool makes. We can also use ChatGPT to field test our essay questions and discussion board prompts to see the types of responses we may get from our students. Those are just a few ideas. We’re really at the beginning stages of ChatGPT’s impact on the educational community. With time, creative educators will find additional ways that ChatGPT can support the work they do. For some additional teaching ideas, check out this New York Times article: Don’t Ban ChatGPT in Schools. Teach With It.

For some other ChatGPT related content, check out:

My blog posts on ChatGPT:
https://the8blog.wordpress.com/2022/12/13/being-an-effective-teacher-by-chatgpt/
https://the8blog.wordpress.com/2022/12/20/teaching-in-a-chatgpt-world/

My podcast episode on ChatGPT:
https://scienceinbetween.fireside.fm/119

Top Posts from 2022 – Part 2

Sometime in December 2022, this blog had its 150,000th reader. I don’t know the exact day it happened or who that 150,000th reader was, but I’m humbled by the fact that so many people have come to this blog for my musings and ramblings over the years. If you’re a regular reader (or a subscriber), I appreciate that you take a few minutes out of your busy work week to read through these posts. As I shared when I first started this blog in 2009, this is a labor of love. I don’t do this to self-promote my work or to feed my own ego. I’m working through things, just like most of you. After more than fifty years of life, thirty years of teaching, twenty five years of marriage and more than twenty years of parenthood, I’m still trying to figure things out. And this space helps me do that. Thanks for being here.

Without any further fanfare, here are the top five posts from 2022.

1. Something Controversial: From April 2022, this post draws on Paul Hanstedt’s book, Creating Wicked Students: Designing Courses for a Complex World (2018) and discusses different ways to structure course content.

2. Reading More Closely: I’ve always struggled with getting students to complete course assignments. In this post from February 2022, I discuss different strategies I use to help my students more closely read the texts I assign.

3. Revisiting “A Politicized Space”: A few years ago, I wrote about trying to appear politically neutral to my students. In this post from January 2022, I revisit that earlier post but discuss research that Jose Antonio Bowen shared in his book, Teaching Change: How to Develop Independent Thinkers Using Relationships, Resilience, and Reflection (2021).

4. The Last Class: Endings have always been difficult for me. In this post from November 2022, I discuss how challenging it is to teach the last class of the semester.

5. Revisiting Start with Thanks: Drawing on the book, How the Way We Talk Can Change the Way We Work: Seven Languages for Transformation (Kegan & Lehay, 2001), this post offers a revised look at how we communicate our gratitude to our colleagues.

Top Posts from 2022 – Part 1

Here we are, another new year together. As we step into 2023, I’m constantly reminded how much I value this space. I know that it’s mostly a one-way conversation where I reflect on something in the open. But I know you’re out there, regularly reading this stuff. I look at the statistics and the people who have signed up to receive these musings in their in-box and I’m honored by that. This has been (another) tough year and if this blog has brought you joy or helped you to reflect on stuff, that’s awesome. Because that’s what it does for me, too. Thanks.

So, without much more fanfare, over the next two weeks, I’ll be sharing the top visited posts from 2022. This week features posts six through ten. I’ll share the top five next week.

6. Promoting Personalization: From March 2022, this post draws on Jose Antonio Bowen’s recent book Teaching Change: How to Develop Independent Thinkers Using Relationships, Resilience, and Reflection. In the book, Bowen discusses ways to foster relationships by making the learning environment more personal so we can better connect with our students.

7. Not in Trouble: As the title suggests, I wasn’t reprimanded or anything. But I did have some explaining to do. In this post from September 2022, I lament that I didn’t get a chance to have a conversation with the student who emailed my university administration about the unfairness of one of my assignments.

8. Being an Effective Teacher by ChatGPT: The fact that this post made the top 10 but was only written in December shows how quickly ChatGPT has entered into the public discourse. In this post, artificial intelligence takes a crack at describing effective teaching.

9. Making Sense of Morale: From February 2022, this post discusses the four factors that contribute to faculty motivation and outlines how they may offer some opportunities for addressing the low faculty morale that is prevalent on many college campuses.

10. Having a Word with Wordle: While Wordle was introduced to the world in November 2021, this post was written a few months later (in January 2022) and discusses how the game can inform the work we do as educators.

Teaching in a ChatGPT World

Last week, I posted some responses that ChatGPT had provided about effective teaching. Before discussing the responses, it’s worthwhile to discuss how ChatGPT works. In last week’s post, I explained that ChatGPT is an artificial intelligence-based chat bot, but I didn’t get into how it actually works. So, I asked ChatGPT to explain itself. Here’s what they said:

ChatGPT is a variant of the GPT (Generative Pre-training Transformer) language model developed by OpenAI. It is designed to generate human-like text in real-time in a conversational context, by predicting the next word or phrase based on the input provided. The GPT model is trained on a large dataset of text and uses a transformer architecture to generate text. The transformer architecture is a type of neural network that is particularly well-suited to processing sequential data, such as natural language text. It uses self-attention mechanisms to allow the model to consider the context of the entire input sequence when generating output.

In the case of ChatGPT, the model is specifically trained to generate text in a conversational context, by predicting responses to input prompts. It uses the input prompts to generate context and generate appropriate responses. The model can generate responses in a variety of styles, depending on the training data it was provided and the specific configuration of the model.

To me, the important part is that ChatGPT draws on its existing database and uses predictive analytics to generate the next word or phrase to a response. Looking at the responses in my post last week, it is clear that the database from which ChatGPT draws its responses saw effective teaching as more closely connected “commitment to student success and learning” and “passion for teaching” than content knowledge. That’s probably not that surprising to a lot of people. But I want to draw on that commitment and passion to outline some ways that we as teachers can better navigate this ChatGPT world. Here are some of my initial thoughts:

1. Don’t ban or block ChatGPT. I know that some folks’ natural reactions to innovations like ChatGPT is to try to make it go away. Schools technically have the ability to block the site and could develop policy to ban students from using it. If we look at the history of technological innovations in schools (calculators, spelling and grammar checkers, smartphones, etc.), we’ll see shortsighted solutions that were shown to be unrealistic in the long term. I see ChatGPT in the same way. Sure, it’s been a pretty eventful few weeks (ChatGPT was released in late November!) but schools will figure out better ways to use ChatGPT than banning or blocking it.

2. We can teach with ChatGPT. To better prepare students in this ChatGPT world, we need to teach them (and teachers) to critically examine the responses it gives. Over the last few weeks, we’ve learned that ChatGPT will provide wrong answers, calculate problems incorrectly, make wrong connections, and offer ethically challenging advice. Now, the responses may get better with time (after all, it is learning) but ChatGPT’s responses offer opportunities that we should embrace rather than avoid. We need to bring the tool and its responses into our classrooms and discuss and analyze them.

3. We need to revise our plagiarism policies. Schools are already dealing with a number of student submissions of ChatGPT created assignments. I urge institutions to revisit their policies and revise them to reflect ChatGPT generated work. For example, my institution’s plagiarism policy states:

“Plagiarism is presenting as one’s own work:
1.
A paper or work wholly or partially done by someone else;
2.
A passage copied from another source without giving credit to the author;
3.
A creative idea copied without giving credit to the creator.”

Is ChatGPT a “someone”? Can artificial intelligence technically be considered “an author”? Smarter legal minds will have to weigh in on this, but we will probably need to clarify some of our policy language. Quickly. (A quick side note, I ran a ChatGPT generated through a Plagiarism checker, and it found no issues. But don’t worry. There is a ChatGPT checker available at: https://huggingface.co/openai-detector/ and it flagged all five of the responses I included in last week’s email with a 95% or greater probability as being fake.)

4. We can use ChatGPT to plan. Over the last few weeks, I’ve heard from colleagues who have used ChatGPT to field test their essay questions to see how ChatGPT would craft responses. Another colleague used ChatGPT to create a lesson plan for an upcoming class. In preparation for my spring classes, I asked ChatGPT to offer some recent peer-reviewed publications that I could include in my syllabus to update some of the readings. It offered a bunch.

5. We may have to assess differently. Many teachers use papers as the main form of assessment in their classes. ChatGPT may cause some folks to reconsider those practices. Some of my colleagues are considering in-class writing assignments instead of papers submitted out of class. Others have discussed bringing back oral exams. In the short term, it’s clear that ChatGPT will have us rethinking all forms of student-created products from discussion forum posts to essays to literature reviews.

While we’re navigating the challenges that ChatGPT poses for our roles as teachers, it’s important that we also consider the opportunities the tool can offer. Sure, it will change some of how we how we do things, but as Seth Godin wrote in a recent blog post about ChatGPT, “Technology begins by making old work easier, but then it requires that new work be better.” In the wake of the introduction of ChatGPT, I trust that the educational community will find ways to make our teaching and student learning better.