By Scott Silver-Bonito
Let’s set the scene: your students are about to have an assessment. Maybe it’s an essay test comparing and contrasting two books, an assessment on geometric expressions, or a Spanish quiz where students need to detail what you can do in different rooms of the house and describe those rooms in writing.
Assessments are a cornerstone of measuring student achievement. As a teacher, of course you want to give your students the best chance to adequately prepare for these assessments and demonstrate their learning. Therefore, you write them a study guide. But what does this study guide look like?
If it’s anything like the study guides of old, it’s a packet with practice prompts and problems, or areas where students restate information they’ve learned in class. I remember for one of my Spanish assessments, our teacher gave us a fill-in-the-blank packet to help us memorize a laundry list of facts from prehistoric to modern-day Spain. I don’t think I’ll ever forget that an ánfora is an ancient vessel that held water.
While well-intentioned, these packet-oriented study guides may not always prepare students for more contemporary assessments in which teachers are asking students to work through real world problems, or solve tasks affecting real world audiences.
I’ve found that students can struggle to find good study techniques and may turn up empty when looking for a study buddy (shout out to my mom for helping me through every Spanish vocabulary quiz ever, even though she took French). Additionally, students may study in a way that reinforces incorrect information or skills that need feedback for growth.
Packets or even lists of practice activities can miss the mark here: they focus too much on a finite set of practice prompts or memorized information, and they don’t build in room for accuracy checks or feedback. Additionally, as a teacher of upwards of 50–100 (or more) students, it’s hard to find the time to check all the student work in progress.
Enter AI.
Using AI for Interactive Study Guides
This year, I transitioned from packets and hyperdocs (a Google Doc with links to practice websites and activities) to AI bots to help students practice their skills and knowledge.
The feedback was clear: students preferred this approach. It gave them timely feedback, and their participation in the study activities increased! Students mentioned in their reflections how helpful these activities were and requested more of them. What started as a pilot a year ago has become a common practice for me as I create my student study guides.
So, how can you start to use AI to prepare study spaces for students? I’ll give you an example of what I’ve done and then give you tips on how to create your own.
My world language students had a quiz on reflexive verbs and getting ready to go on a trip. They had to use vocabulary associated with getting ready and use reflexive verbs in their responses. The format of the quiz was an article that students would read and answer questions about, using the vocabulary and reflexive verbs. Lastly, they would need to respond to a series of short-answer questions, correctly using reflexive verbs to describe their own routines.
While I did give them the typical hyperdoc to study, I reserved two sections for AI study spaces using the School AI platform. A quick note on School AI: it’s a free and paid website that has a lot of teacher-facing and student-facing tools that leverage AI. I highly recommend it, and even the free version is quite well-outfitted.
One caution: While I have found great success with student-facing AI bots, AI is still a developing technology and there are times that AI may produce inaccurate information, so it’s always best to preview the tool with students with the caveats that when in doubt, check with the teacher. Additionally, AI is still lagging in the area of producing visual math models, from the feedback I’ve received and the literature I have read.
AI won’t always be the best answer, nor even a good answer, but it does have spaces where it can free up teacher time and focus to interact with students and more quickly analyze student work and progress. Before we dive into AI study spaces, my advice: start small, be persistent and patient, and know that sometimes AI isn’t the best solution.
AI-generated image by the author
Creating an AI Study Space
Within School AI, I used what they call “Spaces,” which is an instructor-programmed, interactive AI space for students to work. To program this AI study space, I used AI prompting, which is writing guidelines under which the AI bot will act. Side tip: the more detailed you make the prompt and the more you trial and adjust your prompts, the better the AI bot will be for you and your students.
Here is the detailed prompt I used for the study space, which provided students with an area to practice short-answer questions:
- “As a master 7th-grade Spanish teacher, who teaches students taking a 2nd year of Spanish, ask the students short-answer questions, in Spanish, using the attached vocabulary. Please give them feedback, in English, on any grammar mistakes, spelling errors, or sentence structure and have them re-answer the question. The feedback on errors should be in ENGLISH, please. ONLY use reflexive verbs and clothing vocabulary, NOT travel vocabulary, from the vocabulary list, please. Do not use reflexive verbs with pronouns attached. The students don’t know that yet. Student answers should be one sentence in length before moving on to the next question.”
Within this prompt, I gave the AI bot a role: “a master 7th-grade Spanish teacher who teaches students taking Spanish for a 2nd year.” I wrote the activity: “ask the students short-answer questions using the attached vocabulary” (which I attached in PDF form). I then gave the bot parameters on what to fix and also what not to do: fix grammar, spelling, and sentence structure, while not suggesting reflexive verbs with attached pronouns. I also told the bot what answers should look like: “one sentence in length using reflexive verbs and clothing vocabulary.”
When you prompt the AI, make sure to give the bot a role, a task, guidelines, and “avoids”/ “do not dos”. Once you finish a prompt, School AI allows users to test it. I tested the prompt, using completely correct answers and also answers with errors.
The first question the bot gave me was: ¿Qué te pones cuando hace frío? (What do you put on when it is cold out?). I answered using “Me pone” instead of “Me pongo” (I put on…) and the bot immediately gave me feedback that I should be using “Me pongo” and asked me to rewrite my answer. This is a huge benefit to students because it gives them immediate feedback on their written work and an opportunity to correct it without errors – something that would greatly tax me as a teacher if all 98 of my students turned in analog, written work with these practice prompts.
Then I wondered…what if my answer was grammatically correct but didn’t make sense in the context of the question? How would the bot respond? So I wrote: “I put on shorts when it is cold out” (in Spanish) and the bot wrote back, “Shorts are something people usually wear when it’s hot out, not cold. Please rewrite with clothing you’d put on when it’s cold.
Launching the Bot and Benefits
So I programmed the bot, tested it out, and I was now ready to launch it! I activated the bot space and put the link on the students’ study guide.
An added benefit of School AI is that it gives the teacher an “at a glance” view of each student’s progress. You can see a complete transcript of the student’s interaction with the bot, and School AI also gives a summarizing statement about the student’s progress.
For example, in the bot I programmed where students were asked questions about how they get ready, some of the statements I received were: “Student seeks more vocabulary practice; showing initiative,” “Consistent progress in sentence structure and grammar,” and “Engaged in daily routine vocabulary; needs encouragement.”
School AI also flags troublesome student statements (I had one student mention his appendix was exploding, and that was flagged with “Student mentioned bodily harmed”).
An additional benefit to all of this was that students were able to practice a nearly unlimited amount of short-answer writing with unique questions. It takes the onus off of the teacher to create and recreate more practice sentences for students. For math teachers, it can create a large reservoir of word problems.
Since I piloted this first effort at using the AI bot to create study opportunities for students, the feedback from my 7th graders has been hugely positive, and it’s a practice I intend to continue.
The Future of AI in Education
The age of AI is here to stay. I recognize that there is a lot of trepidation around AI and student interaction. We worry that students will misuse it; there are stories of cognitive atrophy, and I agree, if we aren’t vigilant, AI can supplant critical opportunities for our students to construct their own knowledge and learning skills.
However, AI also presents an opportunity for us to reframe educational experiences that have since long been outmoded, such as the classic, informational study guide with all its shortcomings.
School AI presented me with an opportunity to build study spaces that are interactive, informative, and responsive; study guides that take the burden off of the teacher to create and respond to study materials that are student consumables and promote personalized learning without increasing educator workload.
AI study spaces are a powerful example of how we can positively leverage AI to shift the paradigm of education and use AI tools to transform educational experiences for our students.
Dr. Scott Silver-Bonito is a Technology Integration Specialist and 7th grade Spanish teacher in the Wilton (CT) Public Schools. Scott holds master’s degrees in Curriculum and Instruction and Spanish-language literature, a sixth-year degree in Educational Administration and Leadership from Southern Connecticut State University, and a doctorate from Northeastern University in curriculum, instruction, leadership, and learning with a focus on social justice. His special interests include design thinking and adult education.
