I’m teaching myself Spanish. Very poorly, I might add, if my goal is to be a facile speaker and listener. These days I rush through Duolingo, trying to keep my daily appointment with Duo to under 15 minutes. My motivation is low. It was higher six months ago when I was also listening to several Duolingo podcasts and then going through the transcripts. But I got lazy. I’ve also avoided being in a position where fluency in speaking and listening in Spanish matters. I don’t have the skills. My learning efforts are minimal. And what I am good at is recognizing the tricks to speed my way through Duolingo. My reading of Spanish has improved, and my vocabulary is a little wider, but that’s about it.
I will know that I have learned Spanish adequately when I can hold my own in an everyday conversation (both speaking and listening) with another adult without long pauses or being extra slow. It’s a competency that is easy for me to self-assess. My interlocutor will also know whether or not I have the basic competency from a short 15-minute conversation. My goal in Spanish is skill-based. Have I mastered the skill to the desired level or not? It is quite easily demonstrable. Right now, my skill level likely matches a seven-year old, with the vocabulary of an eleven-year old.
What is the skill of Chemistry? Is it analogous to Spanish? How do I know if I have mastered the skill to an adequate level? As a teacher, how do I know if my students have done so? Let’s take first-year college General Chemistry as an example. There is a list of learning goals and student learning outcomes, along with (partially vague) statements beginning with: “By the end of this course, you will be able to…” And how will we know if a student has met the objectives? There’s an assessment. Most commonly it’s a final written exam – which, in my opinion, is a good way to assess the knowledge the student has acquired, both conceptual and procedural.
The word ‘skill’ is more easily linked to procedural knowledge. In an assessment, we can test it by giving students a problem to solve. If the student can solve the problem, procedural knowledge has been demonstrated. In a well-designed problem, conceptual knowledge can also be indirectly assessed. But I have come across many instances where a student can solve a numerical problem with just a vague notion of key conceptual pillars. That’s why my exam problems regularly ask students to explain, explain, explain. It’s why in class or in my office, when a student answers a question, I always follow up with “Why?” regardless of whether the original answer is right or wrong.
I believe that asking students to generate an explanation is crucial. That’s why I hardly use multiple or fill-in-the-blank questions – asking the student to recognize a conceptually sound answer compared to an erroneous one isn’t enough. It is possible to design very good two-stage multiple-choice questions with very plausible sounding conceptual answers, but this takes a lot of work. And I’m lazy. Or I should say, the cost-benefit ratio isn’t favorable at the moment. I’ve tried asking ChatGPT to generate high-quality assessments in this vein, but so far it hasn’t performed to expectations. (It can generate decent open-ended questions with appropriate prompts.)
The move to online homework systems in college-level General Chemistry is practically complete. It’s very difficult to go back and ask instructors to generate and grade homework and problem sets. We have large classes. We have time constraints. We have a host of other commitments. We’re lazy. We recognize the limitations of these systems, and the tech-companies are trying to oblige by convincing us that they are overcoming these limitations. “Adaptive learning” systems, previously a curiosity, is now embedded by all major players. “We’ll take care of it for you,” they say. “We’ll give them the skills and by demonstrating those skills, we know they’ve achieved the learning goals YOU, the instructor, have set.”
How does one assess such skills?
By atomizing them. (Doesn’t this sound soooo
appropriate in chemistry?) There is some value to this approach. If
students achieve fluency on an atomized task, they can “chunk” it into
long-term memory, thus freeing up cognitive resources for more complex and
deeper learning. If that’s how you’re using the online homework system, it’s
appropriate. But don’t make the mistake of thinking that by “mastering” these
atomized skills, the students have learned how to be conversant in the language
of chemistry. Quality feedback is crucial.
But what does it mean to be conversant in chemistry? You have to know the vocabulary. You have to use that vocabulary in the appropriate situation both effectively and intelligibly. Experimenting with ChatGPT has reminded me that a data dump with plausible sounding language isn’t being conversant in chemistry. But it’s hard to nail down exactly what the gestalt experience of understanding chemistry is. That’s because the referents of chemistry are unseen tiny entities, and much of chemistry reasoning is in the realm of abstraction, translated by analogies and metaphors. It’s what makes chemistry particularly challenging to learn: Conversation about the concrete (what you can see and taste) takes place in a seemingly alien world of abstraction such as the fuzziness of orbitals and the fluidity of energy.
Conversation is how we assess fluency. Ideally, I could set up hour-long oral final assessments with each of my students and have a conversation with them by asking them questions. That’s not practical unless I have a very small class. But a written final exam is a reasonable substitute. I have a series of questions that requires the students to be conversant in chemistry by writing down their explanations and supporting their answers with graphs, equations, calculations, analogies, and metaphors. (I have not dared to try a Reverse Final yet.) To prepare them for my final exam, they have to see this being modeled (by me talking and writing things on the board in class) and practice (in homework, quizzes, midterms), and by conversing with me and with each other in the shared language of chemistry.
No comments:
Post a Comment