You are here


teaching, learning, and education

AI Essays: No the Sky is Not Falling

The funniest thing I've heard this morning is teachers talking about using AI to detect AI essays. This, of course, based on the old maxim "Set a 'bot to catch a 'bot." But it is deeply, deeply ironic to use the technology we would forbid our students to use.

People are running around screaming, "Augh! The sky is falling!" But, you know what? People have been saying this about technology forever. I remember when people had an outsized reaction to search engines. "OMG! What if students can just look up the answers to all my questions?" Faculty in my department seriously suggested disabling network connectivity in classrooms to prevent students from having access to the Internet.

Burying your head in the sand is not a solution. I used to say, "If you're asking students questions they can answer using Google, you're asking the wrong questions." And, more recently, I've said, "If you're asking students questions that don't *require* them to start with Google, you're asking the wrong questions."

The thing to focus on is what the technology *can't* do. The AI can come up with human language that's statistically associated with the words you're using, but it doesn't *understand* what the words mean.

Years ago, Randy Phillis and I worked together developing problems for students to solve in class. Using a classroom communication system, you can poll students and then project results, to see whether the students predominantly choose one option or are divided among other options. Some people call these "clicker questions" and think they are asking them to see if students are getting the "right" answers. We realized we could ask students model-based problems and see whether students could reason effectively with the models. This is *not* a skill that the AIs can solve. Yet.

You what else AIs can't do? Anything else. So if you ask students to do a project, the AI can't do that. They can't dig a hole and look for earthworms, for example. Or interpret a set of data. Or build a device.

Students have been able to buy essays for a long time. This is not really all that different.

Writing in Biology: Choosing a theme

Before I started posting about what I do each week in Writing in Biology, I should probably have drafted a post about establishing a "theme" for each semester. The theme ends up affecting many of the exercises I have students do and some of the scheduling of events.

By "theme", I mean a topic or subject area that the course is going to focus on for the semester. This often will affect what organism/object I choose for the first "writing from experience" exercise, what papers I choose for students to look at in week 2, and will limit the scope of what kinds of research proposals and projects we'll work on.

I've tried teaching the class both with and without an organizing theme. My sense has been that it works better to have a theme. A theme usually gives students ideas about the kinds of things they might do. Having no theme leaves students adrift trying to come up with ideas. Students also frequently want to choose inappropriate topics for research (human subjects or vertebrate research that would require complex paperwork to conduct) and choosing a broad theme that excludes those topics, takes them right off the table.

I usually try to pick something that I don't know much about. I do this, in part, because it's an opportunity for me to learn stuff. But also because I'm much less likely to become overly directive if I don't know too much about it. When I've picked a topic I already know pretty well, I find that my opinions end up guiding students too much: they're better off to go into the subject themselves.

Some of the themes worked pretty well. I was particularly pleased with the outcomes of studying vernal pools, tardigrades, garlic mustard, and planarians. Successful themes seem to incorporate a mix of student wonder and importance.

Less successful themes were not disastrous, but didn't pique students' interests for whatever reason. I'm reminded of the semester we studied cockroaches. I picked the theme because I was aware students generally prefer to study animals to plants, but finding an animal to study in November/December is challenging. I thought that we could catch cockroaches in the building and look at demographics/diversity of the populations. But students seemed to regard cockroaches not as "animals" but as mere "vermin" unworthy of study.

One challenge with teaching for participation is that each group of students is different. And even just one or two enthusiastic -- or recalcitrant -- individuals can make a huge difference in the atmosphere of the class. In one semester, I coordinated with the Amherst Tree Warden with the goal of having the data we collected be also a community service learning project with the Town. But I had a student who was seemingly a devotee of Ayn Rand and complained bitterly for weeks about being "forced" to do "free work" for the Town.

In point of fact, when it comes to writing a proposal or project, I don't let the theme interfere with a student, or group, that wants to write about something else. It is always my goal that students who really want to write about any particular thing should be supported in their aspirations. The whole point of my teaching is to liberate students to use the class to pursue their own interests, after all.

Writing in Biology: Second Class

I start off every class by checking in with students to gauge the temperature of how things are going: Any questions? Where are we? What are we doing today? In the first couple of classes, I tend to be more directive with the goal of reassuring students that the class is not some weird, untraditional experience which can rattle some students: In each class, I try to drop the reins of control and give students their head to use the course for their own learning.

In most weeks, I take a few minutes to point out the readings I had asked them to do and encourage them to use the book and other readings as they're doing their writing. I take a few minutes to make a case that this week's reading about paragraphs is particularly important -- for writing the weekly perfect paragraphs and commenting on other people's paragraphs. And to point out that all of their writing should result in carefully focused paragraphs with (referring to the textbook) a topic sentence, consistent order & point of view, cohesive sentences, key terms for continuity, and transitions.

I've also asked them to read "The One Right Way to Talk Science" in Lemke's Talking Science. This chapter, which is actually about avoiding the mystique of science, has an excellent list of characteristics that a writer can deploy to emulate scientific prose. I organize these, with some examples as an introduction to "Uncreative Writing":

1. Be as verbally explicit and universal as possible.
2. Avoid colloquial forms of language.
3. Use technical terms in place of colloquial synonyms.
4. Avoid personification and use of specifically or usually human attributes or qualities.
5. Avoid metaphoric and figurative language.
7. Avoid personalities and reference to individual human beings.
8. Avoid reference to fiction or fantasy.
9. Use causal forms of explanation and avoid narrative and dramatic accounts.

For class, I've provided two bibliographic references that I've asked them to find and skim -- mainly to check that they can find a bibliographic reference with recommendations on how to use the library's proxy service to get access to publications from off-campus. One paper is a review paper and the other is a research paper. My goal is for them to recognize that one has content-based headings and the other has the traditional headings of a research paper. Many of the students have never been aware of the existence of two kinds of scientific articles and it's always interesting to ask them what the authors of the review paper "did". My main point is to draw attention to the structure of a research article: how each section plays a particular role in the manuscript. And how the form is highly synthetic and artificial.

Having been talking for a few minutes, I ask them to do an activity in pairs where I provide each pair a link and ask them to evaluate the linked webpage/site to comments on its reliability and trustworthiness. I provide links to a range of sites including Google, Google Scholar, Wikipedia, Encyclopedia Britannica, Science, Trends in Ecology and Evolution, a predatory journal, news sites, a click-bait propaganda site, think tanks, non-profits, and some parody sites (like the Tree Octopus page). We explore their thinking and then use the CRAAP test as a tool for evaluating reliability.

I then provide an overview of the METHODS project pointing out the goals and various steps and checkpoints along the way. As they're about to start writing a methods section, I have them do a "narrative to exposition" activity I've developed. I ask them create a fine-grained list of all of the (public) activities they did on a recent day, e.g "I woke up. I got up. I put on my glasses. I walked downstairs. I made coffee. I checked my email." After they have the list, I ask them to organize all of the activities into categories. Then I ask them to write a paragraph that explains what each category is and summarizes all of the instances of those activities. The goal is to transform their methods from a story, organized by time, to exposition, organized by the accomplishment of goals, irrespective of order or time. I point out that if they are tempted to use the word "then" in writing their methods, they're probably slipping into narrative rather than exposition.

As the last activity of the day, I show them a picture I've found that is deceptive in terms of scale. It shows an Alfi wood-fired hot-tub, but without any other objects for scale. Most people seem to think it's something for the kitchen, and are astonished when a picture showing a woman bathing reveals its true scale. I use this to point out that pictures they take for their methods projects should include some object for scale, so you can know how big something is: I recommend printing out a page of paper rulers and using one of them, although pointing out that other objects can work as well, e.g. coins.

We end by looking at the prep page for the following week where I've asked them to look through some scientific articles to find a multi-panel scientific figure that is good or has particular qualities they like. And to bring the imagery required for their multi-panel figure, with the goal of working on their figure in class.

Writing in Biology: First Class

Before the first class meeting of Writing in Biology, while the students are coming into the room before class, I invite the students to play a parlor game. I usually show up about 10 minutes ahead of time and, as the students trickle in, I invite them to play Anne Miller from Cripple Creek. In this game, you guess what things Anne Miller likes and doesn't like, usually presented as pairs of things that she likes and doesn't like that, on the surface seem similar or contradictory. For example, Anne Miller likes "glass", but doesn't like "windows". I explain that the same is like science: you need to construct a hypothesis and then test the hypothesis. Once everyone has arrived, we begin the class.

I start out trying to provide students with a reasonably familiar classroom experience. In the long run, I'll want to draw myself out of the center of the class. But I've found that if I start out that way, some students are very uncomfortable, so in the first class, I'm pretty directive -- more than I would ideally want to be.

I introduce myself. I speak briefly about my education and history at the university. I tell them my pronouns, where my office is, when my office hours are, etc. And I tell them that they "have me" for the semester: if they have any questions, they are invited to stop by my office anytime (or make an appointment if they want to be sure I'm there.) And that they're welcome to call on me, not just for questions about the class, but anything else at the University. They're my students and although I might not be able to fix everything, I'm happy to help them figure out who can -- and to help them follow up if something isn't working.

At this point, I ask them to move their chairs into a circle (leaving them in rows to start, is part of the "familiar classroom experience"). And then I have them introduce themselves to the people sitting nearby. I explain that then I'll want each person in the class to be introduced to me by someone else who should tell me their name and the answer to a question. I usually use "what kind of tree would you want to be and where would you grow?" I've occasionally picked other questions, but the tree question works pretty well as a quirky icebreaker. After a few minutes, I start going around the room and have each student introduced to me and begin rehearsing the names in order out loud. "Bob, Karen, George, etc. etc." It turns out it's trivial to memorize 25 names more-or-less in order. I had a professor who could do it with first and last names of 60 students and required everyone in the class to do it as well (each student had to introduce all of the previous students.) I don't do that because it's stressful for the students. Once I get to the end, I usually get a round of applause from the students and I joke that they must always sit in that order from now on. It turns out that many of the students tell me that have never been in a class where a professor learned their names. Which is sad. I promise them that I will do my best to learn everyone's names and they should help me remember.

The key exercise of the day is an observation activity. I give each student an object for them to observe and write about. The object I select varies depending on season and availability. One year, I gave each student a live baby scorpion to observe. In the semester we studied spiders, I gave each student a live Cellar spider (Pholcus). I've often used meal worms (sometimes two different kinds, without telling them that they're not all the same.) Bait shops often have interesting little maggots (spikes) and caterpillars (waxworms) which are good. My favorite are Eristalis larvae called "mousies" in the bait trade. The goal is to have something living and unfamiliar. If something is too familiar (like walnuts), students won't bother to look carefully.

Before class, in the first week's prep page, I have asked them to read Chapter 7 of Louis Agassiz as a Teacher before class. In that story, Agassiz has a student observe a fish for days and write observations. I invite the students to sit for two sessions of 15 minutes to observe and write down everything they notice: To start with just a list of ideas, then to identify questions, and to try to turn the writing into paragraphs. In 15 minutes, we share some observations and reflect a bit on what they've been able to observe and the kinds of tools that might help observation. I've usually printed off some paper rulers that I pass out and encourage them to think quantitatively. After another 15 minutes, we through other observations and then I ask them to count up how many words they wrote in a half hour.

At this point, I go through the structure of the class: I briefly touch on my expectation for weekly draft writing (3 hours divided into 30 minute chunks that they work into their schedule). If they ask how many words they should write, I point out that they just counted how many words they could write in a half hour and that if that was how many word they could write about a maggot, they should surely be able to write about something really interesting. I explain the Perfect Paragraph activity, and describe each of the major projects in turn (Methods, Research Proposals, Research Projects, and Reflective Essay.)

We take a break for a few minutes at this point and reconvene in the Biology Computer Resource Center. I make sure they've set a password for their Biology Account and ask them to confirm they can log into our computers and ask them to log into the Biology Department website I use (which I will describe separately). We spend the rest of the period making sure they can log in, create a blog post as a draft (with some of their observations), create a second post as a Perfect Paragraph, and then comment on someone's Perfect Paragraph.

At the end of class, I point out the prep page for the next week, briefly describe what it contains, when prep-pages will be posted, and ask them if they have any questions.

And that's the first class meeting.

Writing in Biology Retrospective

I first started teaching Writing in Biology in 2002 when several senior faculty retired and the Department was having difficulty finding new instructors to cycle in. In Biology, the course was considered a "short straw" assignment and was taught by a mix of tenure-system and non-tenure-system faculty. I saw potential to have students do projects and write about them and volunteered for the assignment.

Every major at the University is required to offer a junior-year course on Writing in that discipline. The Faculty Senate approved the plan in 1982.

Each department, school or college, in consultation with writing specialists, will determine what kinds of writing competence its majors need.

I don't know if this ever happened in Biology. Two of the retiring faculty gave me their course materials. In consultation with materials from the Writing Program, I developed a version of the class and began teaching it. The course has evolved in small ways since then, but structurally is still largely the same as when I first taught it.

In part, I used the course as a testbed for new technologies: I used wikis early on, then migrated to Drupal with revisions, and most recently have used Google Docs. I even put up an instance of Moodle and used that for one semester, long before the campus adopted Moodle as a learning-management system.

At first, the deliverable for the final project was a manuscript, like the others. But after the chair balked at purchasing a poster printer in response to a new faculty member who wanted one, I made a persuasive pitch that my volunteering to teach the class freed up sufficient funds to justify the purchase so I could use it to have students print posters.

When they shortened the semester by a week, I dropped one of the projects. I used to have students to do a mini "observation project" during the first two weeks of the semester. I really liked the project as it provided a microcosm of what completing one the projects took.

Over the years, I've added a variety of in-class and pre-class activities primarily in response to recognizing specific deficits that students tend to come into the course with: transforming narrative to exposition, writing figure legends, etc.

Over the next few weeks, I plan to write a series off posts to describe the goals of the class and how I try to create an environment for students to achieve those goals.

Busy spring turns to busy summer

Last fall, I proposed to the teaching committee that I would develop an online version of the writing class I teach, as I was aware there were students who would appreciate being able to take the class off-cycle. The department offers four or five sections of the course each semester, but it can still be hard for students to schedule the class, especially if they want to take a semester abroad. I had some reservations—and some in the department were also skeptical—because the University offers a financial incentive for Departments to increase teaching via Continuing and Professional Education (CPE). Students pay for CPE teaching separately—it's not part of their regular tuition—and some departments have been unable to meet students' needs through their "regular" teaching, forcing students to either take classes in the summer or delay graduation. Upon reflection, the teaching committee decided to support my proposal and it moved forward.

It turned out the CPE was also offering small grants to support the transition of classes from face-to-face to on-line instruction. I applied for and was award one of these grants that also required me to participate in an on-line training program for Blackboard. Unfortunately, the awareness of the grant program and my receipt of the award happened so late that I had already made commitments that I knew would interfere with some of the deadlines. And it wasn't clear to me that the "training" would be of any particular help anyway: I've developed on-line classes before and have used Blackboard before. But I figured it might be a useful refresher and the money was not insignificant, so I did it.

I had forgotten how terrible Blackboard is. I mean, I knew, but it's really awful. The interface is horrible. It's clunky and unattractive and inflexible. I remember the first time I used it—we were the first cohort to adopt it—and I reported back to my colleagues how awful it was, they thought I was exaggerating. But when they all had to use it the following semester, they hated it even worse than we had. My strong background in web authoring and technology made it easier for me than for my colleagues. The campus switched to Moodle not long after that and, in spite of an assault on Moodle mounted by a dean, we managed to fight it off for the campus. But CPE remains in the clutches of Blackboard. And it's just as bad as it used to be. Perhaps worse. They have this clunky "portal" bolted on the front now, that you have to interact with if you want to use an avatar and not appear just as a missing-person's placeholder. They want you to to give them information about yourself and populate it with "tiles" that are sort-of like badges. Except no-one does so its rather like a ghost town. And the tiles they have are extremely limited. I pointed out that their "languages" tile didn't list Esperanto which made it unusable for me. The CPE folks submitted a request for me, but Blackboard said they weren't going to add it. But I was impressed they bothered to make the request.

As the semester progressed, I got farther and farther behind. My blog ends near the end of Spring Break when I co-organized NERDSummit. It was great fun and important work, but I did it at the expense of completing the Blackboard training. Then I got talked into going to Libreplanet, which was fantastic, but put me farther behind. Finally, they told me to finish it by a date certain or I would lose the funding. They might have cut me off anyway, but the class had been listed and was filling up, so they really didn't want to do that. I took a weekend and powered through the training. It was awful.

I mean, it was worse than awful. It was filled with textbook-perfect examples of how not to teach. I felt like someone had put a ring in my nose and was dragging my face around on a book and calling it "teaching". They would have videos of powerpoint lectures that you were forced to leave running on your computer to get "credit". The assessment was a quiz you could take as many times as you like where I knew 85% of the answers before beginning the unit. And where many of the questions turned on trivia: Is the menu entry you click on called "Item" or "Object"? About the only unit where you did something that was related to "teaching" was a tool they had licensed called "voicethread". In their course, they had a typo and called it "voicethreat" which was hysterical because the tool requires you to install Flash. I indicated to them that I had quit installing Flash on any of my computers several years ago and believe it a disservice to encourage students to engage in an unsafe behavior to complete a course. Eventually, I received a dispensation that even if I didn't complete the activity, I could still be considered to have "completed" the training. So I skipped it.

I was starting to get caught up toward the end of the semester, when I came down with a spring cold which put me behind again. They threatened again to cancel my class if I didn't get the materials posted. I had gotten them roughed out, but had been reluctant to actually post them because they are all so inter-dependent that I would make a revision in one and then need to go revise three other documents. But I went ahead finally and posted everything.

I've basically created the course in Libreoffice documents, saved as PDF files, and am only using Blackboard in the most minimal way possible. The goal is to be as platform independent as possible so I can switch to something else—anything else—when feasible.

But this weekend, I've finally had a moment to take a deep breath. On Friday, my chair reviewed the course with me. And today CPE is reviewing the course. When they share their review with me, I'll have about three weeks to make revision before the course materials are visible to the students.

It's going to be interesting to see how the course works as an online class. The timeline is extremely short—only 6 weeks—for what is supposed to be 3-credit class. I'm not convinced students can make themselves write that much. But we'll see. The class is full and has a long waiting list. I set the limit at 15, but I'm still concerned about the amount of time that will required to provide feedback to the students. Two of the weeks, I'll be in St. Croix however, which will make for pleasant surroundings while I teach anyway.

Summer is finally here!

Subscribe to RSS - teaching