Innovation is proud to introduce its newest learning tool: structured AI chat.
We created this feature to empower students to practice conversations and engage with course material outside of class time. Although originally designed for world language learners, our AI chat works beautifully across disciplines, making it a versatile resource for content-based courses as well.
At Innovation, we believe technology should enhance learning in structured, meaningful ways. We call our applications “21st-century learning spaces“—they’re carefully designed to meet educational best practices and support student growth.
Teachers have two powerful ways to use AI chat: hosted or hostless.
The AI’s responses follow strict, teacher-defined parameters.
Hosted Chat: This mirrors Innovation’s original synchronous chat app where teachers facilitated real-time conversations between students. The key difference? Now, students can be paired with AI personas instead of classmates.
Hostless Chat: These are independent, self-paced chat assignments that students can complete on their own. But they’re not free-for-alls—the guardrails are still firmly in place:
The chat transcript is automatically recorded.
Students have a limited number of turns with the AI.
When teachers set up a chat, they set the boundaries of the conversation by limiting its length and defining the AI persona’s role.
Students are always responsible to start the conversation. When the chat starts, the assignment is clear and the AI’s interaction protocols are clearly stated.
The AI persona will keep the student on track even if they attempt to distract it with irrelevant questions.
Here, the student tried to distract by asking about sports and the AI brought the discussion back to task.
Once completed, Innovation provides an app to evaluate the quality of the student’s interactions in the chat.
This summer, I will add an AI grading assistant to help assess student work in a chat.
While this was initially envisioned for language students, its use in teaching content became clear. In this example, we set up these parameters for the AI persona:
By the way, teachers can optionally include “accessories” such as a PDF article or a video for students to review before discussion.
Here is how our chat with the Ai started for a critical discussion of the causes of the French revolution.
Our sample conversation went on like this:
Ever true to the parameters set for it, the AI persists in challenging the student to think more deeply and clearly define their points.
The AI chat feature at Innovation has great potential to enrich assignments and promote critical thinking in content courses and linguistic fluency in language classes. Try it out!
Users may have noted the new AI Dashboard in their control panel at Innovation and the new pricing tiers that reference “AI tokens”. What are AI tokens and what can you do with them?
Innovation is integrating AI into every one of its applications now. It is not only a place to teach and learn, it is now a place to create high quality resources to support teaching and learning!
Subscribers to Innovation now have a certain allotment of tokens per month. A “token” is a fundamental unit of text that large language models (LLMs) use to process and generate language. It can be a word, a part of a word, or even a punctuation mark. When you interact with an AI model, your input (prompts) and the model’s output (responses) are broken down into tokens. The cost of using AI services is directly tied to the number of tokens processed. Generally, the more tokens used, the higher the cost. Because Innovation pays per token used via OpenAI, your monthly token allotment is designed to balance value and cost in a fair, transparent way.
During development, I maintained a logging script to see how many tokens I used to complete the various tasks. Unsurprisingly, essay grading is our most token-intensive task, averaging around 1881 tokens per interaction. Vocabulary List Generation is the least token-intensive, consuming significantly fewer tokens.
Of course, token usage varies widely. Some teachers provide detailed outlines when asking the AI to generate tests or discussion prompts based on video or reading material. Others might use the AI heavily for grading essays or enabling student chat discussions. Your usage will shape how far your tokens go.
At the “pro” tier, you get 100,000 AI tokens / month. This resets every 30 days from the date you have a paid subscription. So what can you do with that? Well, based on my own usage (remember, I teach remotely part-time out of Innovation myself!) …
So, what does 100,000 AI tokens actually look like in practice for a teacher? Well, based on my own usage, it’s quite a lot of creative power at your fingertips! For example, you could grade around 53 essays (that’s right, those token-intensive ones!), or generate over 130 sets of test questions for your classes. Need a quick conversation starter for a foreign language class or a debate prompt? You could generate almost 180 conversations. And if you’re building vocabulary, you’re in luck – you could create an incredible 877 vocabulary lists with that many tokens! It really opens up a world of possibilities for creating high-quality teaching and learning resources.
Need comprehension or analysis questions for a PDF or a video? That can be incredibly time-consuming — but Innovation’s Teaching Assistant is here to help!
Just open the Étude app. Upload your PDF or paste your video embed code, then add it to your AI request configuration. In seconds, you’ll have high-quality questions based on your stimulus, crafted in the language and level of sophistication you choose.
The AI Grading Assistant integrated into Innovation is a powerful tool designed to streamline the assessment of student writing tasks.
With just a click, you can apply one of the pre-installed rubrics or upload and use your own custom rubric. After a brief processing time, you’ll receive a detailed second opinion to help you balance and validate your own evaluation of the student’s work. The AI’s assessment is based both on the selected rubric criteria and on the advanced capabilities of a large generative AI model. As of this writing, Innovation uses GPT-4o for essay scoring, ensuring fast, consistent, and thoughtfully reasoned feedback.
My students have always loved playing Jeopardy! Oh, sorry, trademark issue… I mean “Jeopardy-like trivia games in class”.. 😏
Our game is called “Ventura”.
Innovation has had a fantastic app for generating such games for years now. As part of our integration of AI into our whole system, teachers can now employ our AI teaching assistant in generating Jeopardy games!
Just like for creating test questions, teachers configure the request to OpenAI in the Ventura game.
Use the teaching assistant to generate questions. Use them as-is or edit them. Add images or audio clips!
Holy cow, I remember the old days back in the 1990s when I would use PowerPoint to make a Jeopardy game for review day. It took a really long time to enter all the questions and answers even when I had a template game prepared!
Now I can make a game in 2-3 minutes! The test generator is using one of OpenAI’s contemporary models, so you can rely on the question quality.
Those of us who are teaching remotely are starved for interactive apps that let us engage our students beyond screen sharing! Innovation is constantly adding apps and modifications to meet those needs.
Live Sessions
“Live sessions” are interactive sessions that student “join” through the Innovation platform.
Multiple-choice, short answer, and media activity types can all be transformed into live sessions! Just select Live Session from the Create dropdown by your activity in the course playlist. Click Live Link and copy the URL. Send to students in, for example, the Zoom or Teams chat.
Once they join, the teacher host can present one question at a time and await student responses.
Once students respond, teacher is notified and can debrief by displaying responses anonymously.
During the media live session, the teacher presents a slideshow and periodically opens the system for responses, poses a question, and awaits replies.
Activity Monitoring
During composition writing, grammar activities, short answer, and Etude tasks, the teacher can activate the Monitor app. This is found in the Task dropdown for the activity in the playlist. As students work on the task, instructors can view their progress with a minimal time delay. Read more here.
Teachers can hide the student names and the correct answers so they can share the screen with students as they work.
Innovation prides itself on the flexibility to plug in to any learning management system and to be easily integrated in video-conference remote lessons.
From the course playlist, you can send students a link to an activity by clicking the link icon on the right . Paste the link into the video conferencing chat window.
Send students a link to the assessment debriefing (the student’s assessment and correct answers to the task) using the icon below that.
From within the Live Sessions, the same functionality exists.
How to promote academic integrity in remote learning and in-person classrooms with 1:1 laptops
My interest in devising 21st century learning spaces really took off during the pandemic. The school district I was working in at the time had already moved to get all students in middle and high school Chromebooks and all classrooms had interactive projectors (“Smartboards” at the time). I had the advantage of having two perspectives on this, one as a former IT guy (I was district technology coordinator in a few schools in addition to my full-time teaching role and I was a certified network admin) and one perspective as a teacher. I knew we were just co-opting office productivity software for classroom use and it just was not cutting the mustard. Most notably, in our move to digital learning spaces, we lost some of the guardrails we used to maintain academic integrity.
What I mean by digital learning space, a term I use interchangeably with “21st century learning space”, is a software application hosted on the internet in which students conduct their studies and teachers conduct their lessons. My phrase “maintain academic integrity”, well, that mostly just means it was harder to keep kids from cheating.
This situation has come a long way since that time. Schools use a number of content filters, tracking apps, and screen monitoring software that is quite effective. But there are still gaps and I would put Innovation forward as a remedy to fill some of those gaps. Innovation plugs pretty easily into any LMS via convenient links.
Security Tier 1
The apps at Innovation fall into several functional tiers. Tier 1 entails just recording and reporting student activity on the apps. The Proctor is installed to monitor student activity as they interact with the digital learning space. It logs the following student actions:
started task
left the page
returned to the page
pasted in text
resized window
saved work
Tier 1 security on the short answer has some added records, such as notification when student deletes all of their response and the size of newly saved work compared to the answer it replaced. This was devised in response to a student who used to delete all his work and then claim he needed a retake because the system did not save. 🙄
In addition, the short answer task does not allow some other actions such as right click, spell-check, grammerly, activating dev tools, and the like.
Tier 1 security is applied by default on the Etude, short answer, grammar, world language composition, and media proctor. The media proctor records:
video started
video paused
left page
returned to page
duration engaged with video
At the tier 1 security level, the idea is to record detailed information about student engagement and to provide two things: 1) messaging to students showing what is being recorded and 2) reports for instructors who may or may not wish to take action on what the proctor saw. Just telling students that their actions were suspicious (like pasting in text) can serve to deter some mischief.
Tier 1 security is enhanced by the new Monitor app. This allows teachers to view student work progress on a task in real time (well, there’s a 10 second delay after student saves, but it’s still pretty quick). Monitor is available for short answer, grammar, world language composition, and Etudes. The Monitor displays all students who have saved work to the task. Select a student, and their work is shown. The proctor summary shows how many times students are doing each of the proscribed actions.
The multiple-choice app by default has security tier 2.
Security Tier 2
Tier 2 security is enabled by the teacher on the Master page for a task. the master page is accessed from the course playlist under the Task dropdown. Select “Modify test” from the controls at the top, and check the “High Security” checkbox.
When high security is enabled, the short answer task will close and submit responses if the student gives focus to any other page. The student will be locked out until they are formally re-admitted. Re-admit students from the course playlist using the Task dropdown in the controls on the right of the task.
In addition, short answer and multiple-choice tasks can be locked to certain single-use key codes. Once locked, teachers need to provide each student a different code from the list that was generated in order to allow access to the test. This limits attempts to take the test in situations where students have limited chances.
Further, teachers who need this level of security are encouraged to set time limits on the tasks. This will discourage cheating because it often takes time to look things up. In cases where some students get more time on task, you can set exceptions from the testing modification controls in Utilities. Go to Virtual Classroom and Testing Accommodations.
Tier 2 security can be enhanced by having a proctor with students to prevent accessing other devices. In addition, some schools have screen monitoring software like GoGuardian that can assist in monitoring. Perhaps this would be called “tier 3”?
In my current situation teaching part-time as a retiree remotely, I do find it useful to call on students in remote classes. Keeping students engaged in the lesson in a virtual class is a high priority for my attention during a lesson. This is perhaps moreso than in an in-person situation. I think it’s in the nature of digital devices with their many distractions and also due to the limitations placed on human interaction through these tiny windows!
When I am teaching new vocabulary to my French students, I like to use Innovation’sflashcard app. I use this all the time, especially in my beginner level French classes. The app allows me to execute a number of instructional operations: I can show the word, show the meaning, shuffle the word, save out only those words that are problematic for review of a narrower list, practice from definition to term or from term to definition. It really is very flexible.
Now, Reader, in one online high school I work for, all my lessons are one-on-one. So, using the flashcard app is really easy: I share my screen and conduct the instruction.
But teaching to a remote class, even as small as eight students, offers a challenge to maintaining engagement and attention. Last week, I was trying out a new strategy that turned out to work very well. The instructional context is a group of eight students in an AP French class. I needed to teach vocabulary using direct instruction. Here’s what we did: I showed a new term and pronounced it several times. next, I randomly called on a student to repeat and pronounce. then I showed the word’s meaning, then randomly called a different student to type in the Zoom chat to only me the meaning. This protected them from any embarrassment if they got it wrong, although the exercise is set up to be so easy as to limit that possibility. After the session, I sent them a link to a little quiz. The whole thing took about 15 minutes for ten words.
But I was not really great at calling on all students evenly. Some faces were hidden in the way Zoom displays them, so some students did not get called on as much.
There’s a new application now at Innovation that helps teachers to randomly select the next student to respond. It is installed in two places at present, in the main dashboard on the right and inside the flashcards app.
It’s very simple to use. In the flashcard app, click the “Call on Random” button on the left. On the right will appear a simple form. You type in the names, save them, then just click “Select random student”. Voilà! Your next participant!
The app randomly selects a student from the list and then removes them so they cannot be called again until everyone else has been. You can update the list any time.
Look for the random call app to be installed in a number of other places at the site, such as the improvised dialogue app.
Innovation has always developed in response to authentic, practical instructional needs of students and teachers. In retirement, I am enjoying teaching part-time remotely and this continues to inspire new apps and coding enhancement.
You know, Reader, if you take a good look at what you are using to teach in digital spaces, you may observe like I did that a lot of it is software originally designed for office workers. Word processors, spreadsheets, presentation software and the like: these were made for adults doing largely self-directed work in office work. We are so accustomed to these apps that we hardly realize that they don’t ever quite exactly fit for us in the classroom; that we are always creating modifications and work-arounds to make them work. And we get by…
One of my classes this year is an AP French class down in Texas. My objective was to teach them a new grammar point. During our in-class practice, I needed to be able to monitor their work while they were doing it.
Reader, you may already be familiar with Innovation’s grammar learning app. Students learning world languages benefit from practice transforming and generating utterances from prompts. The app meets this need by providing a digital learning space that is interactive. An algorithmic AI lets students know how close they are to the answer, for example, and the instructor can transform the content into a “live session” in which students participate in real time much like the famous Kahoot! game.
Innovation’s grammar app.
Adolescents can sometimes be distractible. In an in-person classroom, I have reasonable observational capacity to notice and redirect distracted students. In remote teaching, this requires some additional effort. What if I could see the student’s’ progress in real time as they worked?
Screenshot of a “live session”, an interactive space where the teacher can pose prompts and students respond in real time interactivity.
People learning new things can sometimes make mistakes. In an in-person classroom, I can wander the room and peer over students’ shoulders. I can try to catch mistakes as they make them and offer correction in a more immediate way. It’s a shame to have to wait a day or two before addressing writing errors. Immediate feedback is more effective so that the other practice examples go well and inculcate the correct syntax. What if I could peer over everybody’s virtual shoulders while they practiced their new writing skill?
The monitor app is now installed at Innovation’s short answer and world language composition tasks. It allows the instructor to view all of the students currently with any saved work on the task. Click the student name, and the instructor can see their work in real time (well, there’s a ten second lag for technical reasons). This work is refreshed every ten seconds. In the short answer monitor, the number after each name tells how many responses they have saved.
In situations where the teacher may wish to share the screen with the class, they can hide the student names and, for the short answer tasks, hide the correct answers.
The monitor, set up for a short answer task, showing students anonymously when needed.
The way I like to use this is as follows: I use two monitors. Monitor 2 is shared with students. I can set the names to “Anonymous” and share the monitor. I select students at random from time to time to check their progress. I may focus on someone who is behind. I may focus on someone I know needs more support (I can see the names before setting anonymous). In monitor 1, on the Zoom or Teams call, I can use the chat to message students corrections, suggestions, redirections if they appear off task, and so forth.
the monitor app, hiding the correct answers in short answer tasks when needed.
To activate the monitor, scroll to the activity in your dashboard course playlist. You’ll find “Monitor Class” in the task dropdown. Monitor is installed for short answer and composition tasks at present. While you are wandering around the site, why not visit our newly opening shops? You can purchase my own activities, PowerPoints, and DBQs for social studies.
When it comes to grading, scores are often reported on a simple 0-100 scale. But, in many cases, it’s better to adjust those scores to make sure they truly reflect how well a student has mastered the material. This adjustment process is often referred to as normalization, and one common way to do this is through a method called z-score standardization.
What is Z-Score Standardization?
Imagine a group of students who took the same test. Some students might have performed really well, while others might have struggled. If we simply average all the scores and compare them to a fixed passing threshold (like 70%), it wouldn’t be fair to those students who performed well beyond the average. Z-score standardization is a way of adjusting scores so that they fit a more accurate and fair scale.
How it works:
Z-Score Calculation: The z-score tells us how far a student’s score is from the average score, measured in terms of standard deviations (which is a fancy way of saying how spread out the scores are). A positive z-score means the student did better than average, and a negative z-score means the student did worse than average.
The formula for calculating a z-score is:
Adjusting Scores: Once we calculate each student’s z-score, we can adjust their scores to match a more standard scale. This is done by applying the z-score to the mean (average) and standard deviation of the group’s scores. The new score is calculated as:
This formula uses the student’s z-score to adjust the score based on how far it is from the group’s average.
Why Do This?
Fairer Grading: By adjusting for how scores are distributed (e.g., a test with a very easy or very hard question), the scores become fairer, especially when comparing students across different groups or assessments.
Removing Bias: Sometimes, individual test questions are biased or poorly written, affecting how students perform. Z-score standardization helps eliminate that bias by focusing on the overall performance of the group.
Outlier Handling: The method also takes into account “outliers” (e.g., one or two students who either do extremely well or very poorly). These outliers can skew results, so they’re filtered out to make the adjusted scores more reliable.
What Does This Look Like in Practice?
Let’s say a student scores a 90 on a test, but the average score for the class is 75, with a standard deviation of 10. To calculate the z-score for the student, we use the formula:
This means the student’s score is 1.5 standard deviations above the class average.
Next, we use the z-score to adjust the student’s score. If we want to bring the class to a higher standard (let’s say the target mean is 80), we use the formula for adjusting the score:
So, the student’s adjusted score is now 95, reflecting their performance in relation to the class and the new target.
Z-score standardization is often mistaken for “curving” scores, but they are fundamentally different. Curving typically involves adjusting all scores on a test so that the highest score becomes a perfect score, or the average score is raised to a certain target (like 70%). This method can unfairly benefit some students and disadvantage others. In contrast, z-score standardization adjusts individual scores based on how far they are from the class average, ensuring that each student’s performance is evaluated relative to the entire group, not a fixed threshold. By considering the spread of scores (standard deviation) and handling outliers, z-score standardization provides a more accurate reflection of a student’s performance, removing the arbitrary nature of curving and offering a fairer and more statistically sound approach to grading.
Innovation makes it incredibly easy for teachers to adjust and standardize assessment scores with our powerful, user-friendly tool. By using z-score standardization, our app helps teachers fairly align scores to a standard scale, taking into account the unique distribution of each class’s performance. With automatic outlier detection and score adjustments, teachers no longer need to worry about arbitrary curving or biased grading. It’s an efficient, data-driven solution that ensures every student’s performance is evaluated accurately and equitably, all with minimal effort on the teacher’s part.
