University of Central Florida

STARS

UCF Created OER Works UCF Created OER

2023

ChatGPT Assignments to Use in Your Classroom Today

Kevin Yee
University of Central Florida, Kevin.Yee@ucf.edu

Kirby Whittington

Erin Doggette
University of Central Florida, Erin.Doggette@ucf.edu

Laurie Uttich
University of Central Florida, Laurie.Uttich@ucf.edu

Find similar works at: https://stars.library.ucf.edu/oer

University of Central Florida Libraries http://library.ucf.edu

This Book is brought to you for free and open access by the UCF Created OER at STARS. It has been accepted for
inclusion in UCF Created OER Works by an authorized administrator of STARS. For more information, please contact
STARS@ucf.edu.

STARS Citation
Yee, Kevin; Whittington, Kirby; Doggette, Erin; and Uttich, Laurie, "ChatGPT Assignments to Use in Your Classroom
Today" (2023). UCF Created OER Works. 8.
https://stars.library.ucf.edu/oer/8
ChatGPT Assignments to
Use in Your Classroom
Today
ChatGPT Assignments to Use in
Your Classroom Today

Kevin Yee, Kirby Whittington, Erin Doggette, Laurie Uttich

First Edition

FCTL Press
Orlando, Florida
ChatGPT Assignments to Use in Your Classroom Today
by Kevin Yee, Kirby Whittington, Erin Doggette, and
Laurie Uttich

Published by
FCTL Press
Orlando, Florida

This work is licensed under Creative Commons BY-NC-SA 4.0.

You are free to share (copy and redistribute the material in any
medium or format) and adapt (remix, transform, and build upon
the material) under the following terms:
• Attribution – You must give appropriate credit, provide
a link to the license, and indicate if changes were made.
You may do so in any reasonable manner, but not in any
way that suggests the licensor endorses you or your use.
• NonCommercial – You may not use the material for
commercial purposes.
• ShareAlike – If you remix, transform, or build upon the
material, you must distribute your contributions under
the same license as the original.
• No additional restrictions – You may not apply legal
terms or technological measures that legally restrict
others from doing anything the license permits.

Cover design by Laurie Uttich. Some individual elements of the

cover were generated by DALL·E 2.

Copyright  2023 FCTL Press

FIRST EDITION
Printed in the United States of America
To our families.

We are indebted to many people for help with research,

ideation, and gathering examples for this book. The list of
individuals includes our colleague in the UCF Faculty Center for
Teaching & Learning Eric Main, and several colleagues in the
UCF Division of Digital Learning, including Wendy Howard,
Rebecca McNulty, Rohan Jowallah, and Rocco Fazzalari. Over
at UCF Libraries, Lily Dubach lent her expertise in organizing
and storing the many articles about AI that have been generated
in recent months.

There has been a great deal of activity on the UCF campus about
AI, as is surely true of many campuses in 2023, and much of this
activity proved helpful in our effort to craft a useful definition
of AI fluency. We thank the many UCF faculty members and
staff who have joined working groups, task forces, and
committees simply out of interest in AI and higher education.

Finally, we are grateful to the members of our leadership team

at UCF who supported this journey, including Provost Michael
Johnson, Vice Provost Jana Jasinski, Vice Provost Thomas
Cavanagh, and CIO Matt Hall.
Introduction
By now, almost everyone has heard of ChatGPT. Its website
(and later app) crushed records in terms of the briefest length of
time needed to amass 100 million users. As a word generator, it’s
simply astonishing! ChatGPT uses a type of generative Artificial
Intelligence (AI) that not only burst onto the scene with
amazing speed from relative obscurity (at least for those not
already working in the field), but also has continued to evolve
rapidly—and by some indications, its development may even be
speeding up. Certainly, there are hundreds of new AI products
released to the public each week. We are already seeing AI in
Microsoft Office and online coursework Learning Management
Systems (LMSs) such as Instructure’s Canvas, and the presence
of AI in these systems and others will only increase in the
future. Those of us on the layperson and user side of the process
are being told to prepare for constant AI evolution, a future of
AI in almost everything we do, regardless of field or discipline,
and a need to develop skills that make us ready to adapt and
keep adapting to the ever-changing mosaic of AI that awaits us.

How Large Language Models (LLMs) Like ChatGPT Work

It’s worthwhile to pause for a brief explanation of how ChatGPT

works. There are many different types of AI, and several of
them have been part of our everyday lives for years.
Smartphone apps that provide driving directions are powered by
AI, as of course are home assistants (Alexa, etc.) and machine
translation apps that effortlessly convert English into another
language, even signs and printed text as seen through the
phone’s camera, and vice versa. And there are many other such
examples in modern life.

i
ChatGPT and several of its competitors are part of a branch of
AI called “generative” AI, which is a category of software that
generates an output after having learned common patterns and
structures. The category includes not only text but also images
and even video. Those that focus on text are called Large
Language Models (LLMs). LLMs can generate text because they
have absorbed billions or even trillions of pages of text, often
described as having been “trained on” the material. This could
include parts of the Internet, published books, academic articles,
and almost any printed and digital material deemed relevant for
a broad audience. Ultimately, exactly what an LLM has been
trained on remains a black box mystery, as few of the companies
have been forthcoming with details. ChatGPT is so named
because it’s optimized to provide a conversation (“chat”) that
optimizes its generative pre-trained transformer (“GPT”)
training.

LLMs like ChatGPT are essentially word-predictors. Based on all

those prior examples of recorded text, they have a good idea of
the next logical word in any given sentence. Thus, these systems
don’t actually think. They don’t even comprehend the meaning
of their words, leading some scholars to compare LLMs to
parrots—they can mimic speech, but don’t understand what
they are saying. Therefore, everyone from teachers to students
needs to remember that these word predictors are not answer-
generators.

Or to put it more accurately, ChatGPT CAN generate answers,

but it is not always an accurate answer-generator. It will almost
always generate an answer. In the rare cases it refuses, it will
claim to not know about the most recent events or what’s
current on the Internet, or perhaps offer a rationale why it
should not generate an answer for your particular query.
Moreover, it will deliver its answer with verisimilitude and

ii
with absolute certainty. It’s understandable why users might
accept ChatGPT’s explanations and arguments since they are
delivered without the slightest hedging or trace of hesitation.
Yet its answers are not trustworthy. Since it’s not accessing a
database of information known to be true, but merely
generating “plausible next words,” it is inclined to invent (often
called “hallucinate”) facts and details wholesale, and baldly
assert them as if they were true. Fans of the board game
Balderdash will recognize a similarity—like players in
Balderdash, ChatGPT tries to convince its audience that it has
provided true definitions. At the same time, while ChatGPT
should be distrusted when it comes to factual information,
academic citations, and specific quotes, it’s actually quite good at
brainstorming and ideation—in particular when creating lists of
sub-topics or bullets that relate to a given prompt.

Challenges and Opportunities for Education

Amid this landscape, it’s easy for an educator to become lost.

There are no rules to follow during such a disruption. ChatGPT
makes it easy for students to copy/paste multiple-choice
questions from online tests and obtain answers. It makes it
equally simple to get essay ideas, essay outlines, and even entire
essays, giving rise to a temptation to cheat and submit
ChatGPT’s output as one’s own work. Instructors everywhere
are left asking themselves how they should react. Should we try
to block access to ChatGPT? Should we adjust assignments so
ChatGPT can’t help students complete them? Is it even
POSSIBLE to continue with traditional assignments in the era of
AI?

Given the omnipresence of AI even in the present day—let

alone in the workplaces of the future—it seems foolhardy to
delay the inevitable. As with the invention of the inexpensive

iii
calculator in the 1970s, the genie of AI cannot be stuffed back
into its lamp, and educators must learn how to adjust. We need
to learn what skillsets are helpful for students as they prepare
for careers that include AI. We need to develop new mindsets,
both for us and our students, that help them thrive now and in
the future. And most of all, we need concrete ideas about how
to implement these lofty goals. It may sound great in theory to
embrace the new world of AI in our classrooms, but what
exactly does that look like? What kinds of assignments can we
design that meet our intended learning goals yet aren’t prone to
students cheating using AI tools? Answers to these questions
have been in short supply in books published until now
(although in fairness, less than a year has gone by since
ChatGPT went viral, so it’s hardly reasonable to expect a flood
of books). Even in blogs, discussion boards, and published
articles, concrete examples of action to take in the AI era only
appear in fits and dribbles. This volume aims to bridge that gap
by suggesting practical assignments and in-class activities that
create AI fluency in students.

A New AI Mindset for Students and Instructors

Teachers and faculty everywhere first need to adopt a mindset

that acknowledges the availability of AI and the likelihood that
students will use it. As a result, we need to adjust our
expectations of students. With online tests, maybe we should
stop assessing for their recall of basic information. Or, if this is
necessary, as it surely is at some educational levels and in some
disciplines, perhaps we only conduct this kind of assessment in
controlled, authentic environments where the use of AI is both
prohibited and functionally impossible, such as in-person
testing. In the case of out-of-class writing, we might need to
concede that the centuries-old skills of idea-generation and
essay-generation from scratch might now and forevermore be

iv
the realm of artificial intelligence… or at least assisted by AI.
It’s unreasonable to expect students to resist a tool that can do
the exact assignment for them, especially when it’s free, easily
accessible, and difficult for teachers to detect. Thus, we need to
teach students different skills related to using artificial
intelligence, rather than avoiding it.

Students have long had access to assistance. Calculators assist in

math functions. Current websites will create graphs of complex
formulas, while others can assist with finding areas of irregular
shapes. The assistance continuum extends to writing as well.
Brainstorming a writing topic with a peer is a version of
assistance, as is performing peer editing—two activities
universally endorsed by educators. Spellcheck is now an
accepted norm, and increasingly so are virtual grammar
checkers. Both latter examples lead us along the continuum of
human-to-machine generation. Perhaps the time has come for
us to progress further along that continuum and accept that
machine assistance is not just possible but even desirable in the
ideation and initial drafting phases of writing.

AI Fluency

Clearly, students will need new skill sets in the future to meet
the challenges of future workplaces. Much has been
accomplished toward career readiness through the efforts of the
National Association of Colleges and Employers (NACE),
particularly through the definition of eight core competencies:
career and self-development, communication, critical thinking,
equity and inclusion, leadership, professionalism, teamwork,
and technology. While one might conceivably place AI
familiarity within technology, we suspect the AI revolution
might be significant enough to warrant the eventual creation of
another core competency.

v
We view AI fluency as consisting of seven components:

1. Understanding how AI works

2. Deciding when to use AI (and when not to)
3. Valuing AI
4. Applying effective prompt engineering methods
5. Evaluating AI output
6. Adding human value
7. Displaying digital adaptability

These components are, in our view, broad enough to capture AI

fluency for not only ChatGPT and all LLMs, but also extend
beyond generative AI to other types of AI as well.

The first component, understanding AI, is important because

there are different branches of AI—each with its own strengths
and weaknesses—and one must understand the AI currently
being employed to fully grasp its capabilities. LLMs like
ChatGPT, for example, may be prone to hallucinations, but this
is not true of every type of AI. Artificial intelligence tools of the
future may not construct output in the same fashion, so it’s
important to have a minimal understanding of how the AI tool
at hand creates its output.

Deciding when to use AI and when not to is the second

component. An experienced AI user must exercise sound
judgment about the output of a particular AI. With ChatGPT,
we know that it’s neither safe nor ethical to copy its output
wholesale and represent this text as something created by an
individual. There are also ethical issues of ownership and
copyright, including the works of deceased creators. On the
other hand, some uses of AI may be warranted, or even desired.
In some circumstances, instructors may want students to use

vi
ChatGPT to brainstorm ideas, or to assist in creating an
assignment.

Valuing AI, our third component, is borrowed more or less

wholesale from Bloom’s taxonomy in the affective domain. This
domain is concerned with learners’ attitudes and their
appreciation. “Valuing” is the middle section of the affective
hierarchy, with “responding” and “receiving” above it, and
“organization” and “characterization” below it. We include this
component in our own list of necessary components for AI
fluency because the technology represents such a tectonic shift
in our relationship to knowledge and information, and not
everyone initially responds to AI with the same enthusiasm. Yet
to be fluent in AI, one must appreciate its usage.

Because AI doesn’t have the lifetime of experiences a human

does, it is extremely poor at reading between the lines, or
knowing what an imprecisely worded question is actually
asking. Therefore, our fourth component to AI fluency is
creating effective prompts that elicit useful or desirable output.
As the common phrase goes, if you put garbage in, you’ll get
garbage out. We need to think about prompts (the question
posed to the AI) in ways that are systematic, intentional, and
deliberately plotted. While some disciplines already train
students to think with these methods, especially about the
architecture of programming or arguments, many do not.
Prompt engineering is in many ways a discipline unto itself, and
we all need to become better at it.

A truly critical skill, especially with ChatGPT and its

hallucinations, is the ability to analyze and evaluate AI output,
which is our fifth component of AI fluency. We are increasingly
seeing deepfakes in images and videos concerning public figures
and celebrities, such that one truly should not trust one’s eyes

vii
viewing digital images. We know that ChatGPT invents facts,
names, and publications, and it does so with such confidence as
to border on chutzpah. Users need to remember to approach AI
output of all types with appropriate skepticism, a skill we likely
need to develop further.

The sixth component is perhaps the most important of all when

pondering how our students will attain jobs in the future:
adding human value to the AI output. Because AI can already
automate so many tasks—and because future artificial
intelligences will continue removing human agency from
additional processes—the only employees needed in the
workplace of the future are ones who can add additional value
to what the AI creates. This might look like correcting the AI
output or applying/integrating it into other systems and
processes that the AI cannot perform.

Finally, the seventh component is digital adaptability. We

recognize that artificial intelligence will continue to evolve; in
fact, many believe its evolution and advancement will accelerate
over time. As a result, people will not stay fluent with AI if they
are habituated only to the one artificial intelligence system they
know. There will assuredly be future AI products, and these
need to be approached with an attitude of curiosity and
optimism, or at least not with reluctance, irritation, or
resignation that yet another new system needs to be learned.
We will all need the kind of disposition that welcomes lifelong
AI learning and the flexibility to keep our attitudes positive as
we embrace ongoing AI change.

viii
 Practical Assignments

If we adopt the idea that the mindsets of instructors and

students alike must change and accept that some form of AI
fluency is a necessary desirable component of the curriculum,
then their greatest need is assistance with the implementation
of these concepts. Thus, we provide extensive examples of
assignments that can be used in class or as homework. These
ideas for assignments, and the principles that underlie the need
for them, form the heart of this volume, and indeed make up
the vast majority of its contents. Each example starts with the
actual assignment to give to students before also offering
explanations for why the assignment is worthwhile and how to
implement it.

Many of our assignments are designed to show off what

ChatGPT can do, while others are intended to imbue various
features of AI fluency in students. A third group suggests ways
to leverage ChatGPT in writing assignments, recognizing that
we might need to revisit our original reasons for assignment
writing and re-examine the student learning outcomes we are
attempting to achieve. In the future, will “writing” really mean
“editing” rather than “drafting/composing”? Will future careers
involve editing and refining AI output instead of starting
writing, even analysis, from scratch?

ix
 Scope, Reach, and Organization of This Book

The tips and assignment ideas provided in this volume are

specifically customized for ChatGPT, owned by the company
OpenAI (which also owns the popular image generator DALL-
E). Yet many of the tips and ideas contained here can also be
employed in other instances of LLM-type AI, such as Bing Chat,
Google Bard, and Meta’s BlenderBot. For that matter, we
recognize that today’s leaders in LLMs technology may not be
the leaders of tomorrow, or LLMs might not even be the AI that
matters mere years from now.

All of the sample prompts provided in this book were vetted

with ChatGPT 3.5 (the public and free version) in mid-2023 to
verify that they would provide interesting and relevant output
that might be profitably utilized across disciplines. Future
searches of ChatGPT, or of other LLMs, might not yield
productive results. That said, it is our hope and strong suspicion
that many, if not all, of the sample prompts provided here could
apply to other LLMs beyond ChatGPT as well. We expect that
most of these strategies, in other words, could be used by almost
any related AI.

As for AI-generated text within this volume: there isn’t any. We

wrote this book in early and mid-2023 without using AI, except
in limited ways to test sample AI prompts for each of the
assignments. While we recognize that future book-length works
may opt to follow our advice about using AI to help outline and
chart writing projects, our own process only did so as
verification and after-the-fact analyses instead of as first steps.
We find it to be natural that current pedagogy experts and
holders of terminal degrees may continue with their established
composition practices that do not use AI in the initial stages,
while the opposite may become more common for

x
undergraduates in the next few years. Eventually, of course,
these undergraduates will become our institutional colleagues,
and yet another shift in mindset and practices may become
advisable and necessary.

While the book is organized by a contiguous set of numbers, it

is subtly divided into sections without being overly labeled so.
The first several assignments explore prompt engineering and
provide explanations for why we need it as well as how to teach
it through assignments. Then comes a few assignments about
using ChatGPT for searching, since it can do things search
engines cannot. Up next are assignments that train students on
evaluating AI output, as well as a section of suggested
assignments that show how ChatGPT can be used to analyze
text or data. The largest section concerns assignments related to
writing, as might be expected. The penultimate section provides
ideas for using ChatGPT to generate useful output, and the final
section has assignments designed to help students recognize that
ChatGPT can be used to improve their study habits and
practices.

We hope this book will provide you with support during these
exciting—and daunting—times and inspire you to explore the
possibilities of engaging ChatGPT and other AI tools into your
curriculum.

Kevin Yee
Director, UCF Faculty Center for Teaching and Learning

xi
 1
Prompt Engineering:
Conversation
Create an assignment that forces students to
engage ChatGPT in a conversation.

Unlike search engines, ChatGPT will remember previous

prompts and requests from the same session (until the user
clicks on the “new chat” button). Knowing this, students can
both refine search results and save time by intentionally
interacting with the software in a conversational manner. This
involves “talking to” ChatGPT the way you would talk to any
human. We wouldn’t restate the original question if we were
speaking with a human adult and wanted to clarify with a
second sentence; we would just provide the second set of
specifics. For example, if we first asked ChatGPT to identify the
TV show that was a popular thriller/horror show, the software
is likely to identify several possibilities. But the second prompt
doesn’t need to restate the information already digested and
could skip straight to the clarification “I mean the one that has
the Upside Down.” This technique, known as prompt chaining,
can also be part of an intentional strategy to withhold part of
your ultimate ask until the second or third prompt—such a plan
can be especially useful if you are seeking first a summary of a
text, and then an analysis of its strengths and weaknesses.
Asking everything all at once could be confusing for both

1
humans and software, and often it’s better to pull the questions
apart.

While it seems simple enough and ostensibly thorough enough

to just inform students of the possibility of an extended
conversation with ChatGPT, not requiring them to do it
explicitly on an assignment creates the risk that they will forget
to attempt it later. Practice makes perfect, as the expression
goes, so it seems pedagogically sound to actually force students
to practice. One way to do this is to give them the first prompt
for ChatGPT, and then ask them to read the AI output carefully
and ask at least two follow-up questions for clarification.

Sample ChatGPT prompts:

1. “Explain string theory using everyday words.” / “What

do you mean the strings vibrate?” / “What’s another
metaphor to help explain how string theory works at
both large and small scales?”
2. “Define intersectionality and its origins as a theory.” /
“How have modern theorists gone in directions different
from Crenshaw?”
3. “What are some of the better-known theories of justice
in philosophy?” / “Say more about John Rawls’s theory.”
/ “Explain the Original Position using a different
metaphor or simile.”

2
 2
Prompt Engineering:
Regenerate a Response
Ask students to generate multiple responses from
a single prompt.

When asked to run a search in search engines, the engine will

reproduce the same search nearly every time. This is a good
thing when you are looking to remember what came up for that
search you had previously conducted, but after a few times,
you’ve browsed many of the possibilities. On the other hand,
ChatGPT allows you to regenerate a response based on the same
prompt for a different result.

After the AI tool has completed a response to a prompt, the

“pause response” button will change to a “regenerate response”
button. This button will allow you to ask ChatGPT to write a
different response to the prompt you already entered. This
regenerated response will use the same information to rewrite
the sentences in a different pattern, order, or tone. This is one of
the features that makes AI detection difficult since a unique
output will be given even if the prompts are identical. In
ChatGPT a regenerated response will appear where the first
response was posted. You can toggle between the two generated
posts by clicking the number next to the top of the AI-
generated response.

3
Regenerating a response can prove useful in conjunction with
several other tips in this guide. For example, you might have
students ask ChatGPT the same response multiple times and
compare the different results. You might ask students which
response is most accurate or persuasive. You might also ask for
students to regenerate a response to see if it pulls from the same
or different sources. You can then have students compare the
different sources for the regenerated responses to see which are
more reliable.

Similarly, you can ask the AI software to regenerate a response

after a given amount of time. While ChatGPT 3.5 is not able to
gather new information from the internet, users can provide
feedback as to whether the information provided was accurate.
For this reason, waiting before asking the AI to regenerate a
response can provide different information than it did
previously. Depending on how others have been “training” the
AI, the information could be more accurate or less accurate. It
might be interesting to try this both with prompts that the AI
answered incorrectly or correctly to see if there is a change.

By asking students to consider the different responses provided

by the same prompt, you can begin a discussion about the
strengths, limits, and ethics of using AI. Questions can arise as
to whether the outputs are comparable in terms of their
substance, accuracy, tone, bias, etc.

Sample ChatGPT prompts:

1. “Explain the way to find the area under a curve on a

graph.”
2. “Compare and contrast the work of Claude Monet and
Berthe Morisot.”

4
 3
Prompt Engineering:
Rephrase Prompts
Require students to ask ChatGPT a detailed
question in more than one way to refine the
results.

Much like you might ask a person to rephrase a question or a

sentence if you do not initially understand, you can do the same
with AI. Generative AIs look for the most likely answer to your
prompts. Therefore, it will look at the likelihood of what should
come next and select what has the highest probability. But we
are not always asking about the most likely response to our
question.

As previously discussed, ChatGPT will provide a different

response every time it is asked a prompt. This is the nature of
the generative aspect of AI. While you are able to get a
multitude of different responses to the same prompt by asking
ChatGPT to regenerate a response, you can also rephrase your
prompt to get more tailored responses. Regenerating the
response allows the user to see the same information
represented in a different way. Rephrasing will cause ChatGPT
to potentially generate a different answer or a more nuanced
answer, because it does recognize that this is the same prompt
asked in a different way.

5
Prompt rephrasing is a form of having a conversation with the
AI, except instead of altering the question to get new
information you alter the prompt to get information more
relevant to what you were searching for in the first place.

Sample ChatGPT prompts:

1. “Describe the mechanisms through which muscles

contract.” / “How do muscles contract?” / “What makes
our muscles contract?”
2. “What is forecasting?” / “What are the different methods
used for forecasting?” / “What are the benefits of
different forecasting methods?”
3. “What are the major points in history that led to
Guatemala’s independence?” / “How did Guatemala win
its independence?” / “Who were the important figures
that helped Guatemala win its independence?”

6
 4
Prompt Engineering: Context
and Specific Requests
Teach students via “graduated” prompts how to be
specific and detailed to maximize ChatGPT
output.

Search engines will look for keywords in websites or articles to

match the search criteria you prompted. For this reason, the
search engine may or may not provide new information based
on the level of detail you put into the search engine. ChatGPT
will tailor responses based on the level of detail you provide in
the prompt. The more detail provided, the more nuanced the
response will be. Much like a conversation with ChatGPT,
providing increasingly more detailed prompts will allow
students to get an answer that is more accurate or useful for
their purposes.

Given the massive amounts of data that the AI model was

trained on, there may be several similarly named or overlapping
concepts. We can see this with general search engines when we
put in our own name. You are able to see how many different
people have the same name. Giving more information such as a
state, city, or job title, can narrow down the number of returns
to ones that would be more relevant. This same logic occurs
when using ChatGPT.

7
Students should be provided with an understanding of how the
prompt they put in will impact the response given by ChatGPT.
The broader the prompt the more broad—and possibly
inaccurate—the response will be. Begin by having students
input a very general inquiry. This could be just the name of a
famous person or a piece of machinery used in the industry.
Question students about the output from the AI. Is it useful?
What type of information does it give? Afterward, ask students
to add more details to their prompt, such as a specific date,
accomplishment, or event, and see what the AI generates. You
can then have students compare and contrast the two responses.

Students might also use the information they find in the broader
prompt to provide more details in their subsequent prompts.
This could be useful when students are just introduced to a topic
or do not have a lot of background information to create a
detailed first prompt.

Sample ChatGPT prompts:

1. “What is the difference between APR and APY? / When

would you use APY vs. APR?”
2. “What are the different types of joints that can be used
when constructing a boat?” / “What types of joints can
be made with just pieces of wood?” / “What is the
difference between a tongue and groove joint vs. a
rabbet joint?”
3. “Why is shutter speed important?” / “What happens
when you have a higher shutter speed?” / “When would
you want to use a slower shutter speed?”

8
 5
Prompt Engineering: Tone
Make an assignment for students to obtain
ChatGPT results that match a certain tone.

If not given specific instructions on HOW to write, ChatGPT

will return results in a neutral tone that suggests impartiality
and conveys quiet assuredness. Its default tone seems definitive,
worldly, and devoid of doubt. However, the output’s tone can
be changed upon request. You can ask ChatGPT to answer in a
playful tone or a sarcastic one. Challenge it to answer a specific
prompt with the passive-aggressiveness of a petulant teenager,
for example, or to adopt a particularly strict tone of unamused
authority. By exposing students to the breadth of possible styles
in ChatGPT’s answers, we allow them to better see the
possibilities for their own ongoing prompts of ChatGPT. Plus,
the very novelty of an artificial tone on a serious subject may
cause students to pay closer attention than would otherwise and
maximize their learning in the process.

One assignment idea might be to ask students to generate the

same response three or more times while specifying that the
output should be delivered in a different tone each time. You
might try creating a list of possible tones, from which the
students can make their individual selections.

9
Sample ChatGPT prompts:

1. “Write a summary of the voyage of Odysseus in an epic

tone.”
2. “Describe the process of photosynthesis in cacti in a
whimsical tone.”
3. “Describe the work of Stephen King in an ominous
tone.”

10
 6
Prompt Engineering:
Specified Style
Ask students to exercise their creativity by asking
ChatGPT for explanations or summaries delivered
“in the style” of someone famous.

Since ChatGPT was trained on a large number of books, articles,

blogs, poems, lyrics, and many other forms of writing, it can not
only recognize but even replicate the style of many known
authors and artists. Asking students to generate an explanation
of a scientific principle is one thing, but it’s an entirely new
level of fun—and thus permanence in their memory—if the
message is delivered with a layer of pop culture they might
recognize.

Rather than require students to adhere to a strict set of prompts,

authors, and artists you have pre-selected (at least some of
whom they may not know), it may be more effective to let
students brainstorm their own authors and artists to use as
inspiration. This has the added benefit that their personalized
contribution generates, by its very nature, more investment and
thus ownership in the eventual answer, as well as basically
guaranteeing the chance that they will recognize the
author/artist and be tickled by the unexpected juxtaposition.

11
Learning that incorporates fun has the best chance of becoming
internalized and therefore permanent.

Sample ChatGPT prompts:

1. “Write a sonnet in the style of Shakespeare that explains

the third law of thermodynamics.”
2. “Rewrite Rihanna’s song “Umbrella” using language
from Tolkien, but make the song about tort reform
efforts over the years.”
3. “Using words and rhymes like Dr. Seuss, explain the
Missouri Compromise in five paragraphs or less.”

12
 7
Prompt Engineering:
Sophistication
Encourage students to include in their prompts
how sophisticated of an answer they want.

When not given any specific instructions on how to respond,

ChatGPT will provide output at a high school level, a decision
made to maximize its utility to audiences of various educational
backgrounds and provide a default voice that is, hopefully,
accessible to the widest possible audience. However, this is a
feature that can be changed upon request. ChatGPT can increase
its level of sophistication and presumed background knowledge
of the user when asked to answer like a grad student or
professor in that field. Alternately, it can answer with the
diction and sophistication of a grade-schooler, if desired.

College-aged students may well desire a professor-like answer to

a question they are tackling for school since they’ll be expected
to provide details and explanations on par with that level of
sophistication. At the other extreme, however, college students
may find value in asking ChatGPT to provide answers optimized
for someone younger, particularly for theories and ideas that are
conceptually very difficult.

One suggestion to encourage students to try this out on their

own is to require a low-stakes assignment (such as a discussion

13
board post) where they post their experiment. Ask them to
prompt ChatGPT to explain a concept from your class by
invoking a certain level of sophistication, then post both their
prompt and ChatGPT’s response in the discussion board.

Sample ChatGPT prompts:

1. “Explain red shifting to me like I’m in second grade.”

2. “Summarize Judith Butler’s argument about
performativity with the sophistication of a professor in
the humanities.”
3. “Define the principle of Mutually Assured Destruction
using examples from playground interactions among
grade schoolers.”

14
 8
Prompt Engineering: Length
Ask students to alter the prompt to provide the
same information in different lengths.

With the massive amount of data that ChatGPT pulls from,

there is no standard length to a ChatGPT response. AI will
attempt to provide a comprehensive output to answer the
prompt. This is usually in the form of several paragraphs,
typically around the five-paragraph essay format. While this can
be very useful, it can be time-consuming to read, provide extra
information that is not necessary to the prompt, or simply not
be relevant for your needs. For example, many of the responses
will start with an introduction to the prompt and end with a
conclusion paragraph that summarizes the previous points. If
this type of output is not what you are looking for, specifying
the length of the response (one paragraph, five-paragraph essay,
etc.) or the format of the response (bullets, steps, etc.) can assist.

When users first start using ChatGPT, the idea of providing

more information is useful. It allows students to get more
information than they would by simply typing their query into
a search engine. But in certain situations, it is helpful to get a
clearer, paired-down, or concise answer. Varying the length
allows ChatGPT to provide you with more or less information
based on your prompt specifications. These specifications
provide varying amounts of detail. Shorter lengths will require

15
ChatGPT to summarize or leave out certain details. Asking for
longer lengths is more likely to provide erroneous details.

There are several ways to let ChatGPT know how long you
want your response to be. The first is writing it as part of the
prompt you put in. For example, “summarize cellular respiration
in five bullet points” or “In one paragraph discuss the role of
mitochondria”. You can also use parenthesis to denote the
length at the end of the prompt. Each of these will be
considered when ChatGPT generates its response. Students can
use the various links to fit the purpose of different assignments
or compare the outputs based on their accuracy, detail, or
generalization.

Sample ChatGPT prompts:

1. “Give me a two-sentence summary of the importance of

superconductors.” / “Provide a two-paragraph summary
of why superconductors are important.”
2. “Write a short story about a cat in a coffee shop.” /
“Write a four-sentence story about a cat in a coffee
shop.”
3. “In a five-paragraph essay format, write the difference
between constructivism and behaviorism.” / “In three
bullet points, write the difference between
constructivism and behaviorism.”

16
 9
Prompt Engineering: New Chat
Create an assignment where students see the value
in restarting chat.

ChatGPT remembers previous prompts and its own subsequent

outputs, essentially having a conversation with the user over the
course of several prompts. When this connection to the prior
conversation is desired, it creates a useful output. Sometimes,
however, the user is actually asking a new question.
Unfortunately, unless ChatGPT is told this is a new, unrelated
direction, the output is often colored by the earlier prompts in
an unwelcome fashion. To obtain a “pure” output to a new
question, it’s advisable to press the “new chat” button before any
change of topic.

The interface makes it easy and intuitive to continue asking new

questions without thinking to reset the chat history. As a result,
students need assignments that allow them to experience this
functionality. One suggestion is to require them to first ask a
series of prompts without re-setting the chat and then end with
a prompt that is related to the first one yet colored by the
conversation. They should paste this to a discussion board post,
then return to ChatGPT, click the “new chat” button, and input
the final prompt again. This second output should be pasted
below the first one in the discussion board, and then they can
analyze the differences.

17
Conversely, students should be encouraged to save relevant
conversations with ChatGPT and other AI tools. When a
prompt is first typed into the box, ChatGPT starts a thread from
which ChatGPT can pull or reuse previous information or
questions asked. By default, ChatGPT saves these chats in the
column on the left. Students can rename the chats by clicking
on the pencil icon. ShareGPT, a Chrome extension, captures the
full conversation with ChatGPT and generates a URL, allowing
students to share the conversation.

18
 10
Search: Narrowing
Search Results
Teach students how to purposefully narrow the
responses ChatGPT generates.

At times we know exactly what we are looking for when we

begin our search. In these instances, the use of quotation marks
and Boolean operators can assist AI tools in returning exactly
what we would like. Quotation marks signify phrases that must
be presented in the results exactly as they are written. Boolean
operators such as AND, OR, and NOT indicate which words,
phrases, or topics that should or should not be included in the
results. Using these in the original prompt can reduce the
amount of back and forth often needed when using ChatGPT.

Other times, we may not know exactly what we are looking for
when we begin our conversation. For these instances, the
conversation with ChatGPT may point us to what we want to
know more about, and we can then add Boolean operators or
quotation marks to our original prompts and ask ChatGPT to
generate a response again.

Additionally, we can show students how to prompt ChatGPT to

elaborate on a given point that it has raised by using the
language and acronyms ChatGPT created in its first response.

19
Since the ChatGPT will “remember” the thread, it will answer
the prompt as if it is a continuation of the conversation.

Sample ChatGPT prompts:

1. “What are career opportunities for computer science OR

software engineering OR computer engineering (and
NOT programming) AND salaries?” / “Why do software
architects earn more?” / “What do I need to do to get a
job in that field?” / “Name 10 places that hire.”
2. “What is the role of metrology in the optics and
photonics field?” / “Explain these ‘stringent performance
standards.’” / “Which tools measure ‘colorimetry?’”/
“Name fields that use ‘spectroradiometers’ (NOT
‘metrology’ OR ‘photonics’ OR ‘optics.’

20
 11
Search: Explanations for
Wrong Answers
Direct students to look up answers they got wrong
on an assignment or assessment.

Prior to the advent of AI tools, the onus of describing why

certain answers were incorrect fell on the instructor. With the
use of Internet searches, students could sift through web pages
to see if something matched what they were looking for and
then evaluate it for accuracy. Because the students answered the
problem incorrectly the first time, it was unlikely they would be
able to easily determine why their answer was wrong without
the instructor’s guidance.

If unresolved, students may retain their misconceptions,

impeding their ability to acquire further knowledge on it later
and hindering their future learning. ChatGPT offers a means
through which either teachers or students can ask questions and
search for answers to common sticking points. For students,
they can also use several of the prompt engineering techniques
to help them better understand what ChatGPT generates.

As part of their assignments or to earn lost points on an

assessment, instructors can have students look up the correct
answers to problems using ChatGPT.

21
Sample ChatGPT prompts:

1. “I was asked, ‘If the wage rate decreases by 8% and the

quantity of labor demanded increases by 4%, what is the
labor elasticity?’ I responded, ‘.5.’ Why is that wrong?” /
“Can you provide more details on a step-by-step basis?” /
“Generate 10 similar questions so I can practice.” /
“Show answers and provide detailed explanations.”
2. “I said, ‘The American Revolution and the French
Revolution were similar because they were both
influenced by Enlightenment ideals, social inequality,
and taxation.’ Why is that wrong?” / “Are you saying the
French Revolution was more violent?” / “When you say,
‘The American Revolution maintained a degree of
respect for legal processes and the rule of law’ while the
French Revolution did not, it sounds as if this is an
‘American’ argument and perspective. Is it? Why or why
not?”

22
 12
Search: Seminal Scholars,
Studies, and Other Materials
Have students report what the AI generates as
seminal scholars, articles, or books within a given
field or topic.

Learning a new topic can be daunting. There are many dates,

people, topics, theories, laws, equations, etc. to be remembered,
and perhaps retrieved at a later time for other purposes. For
many of us, it may be difficult to remember things we just heard
about, or it could be challenging to remember things we learned
previously. Often, we might be left with only a vague idea of
what we want to remember. Search engines have allowed us to
search for queries of which we only have a partial recollection.
For example, we could type in “What is the book with the
caterpillar that was still hungry”? The search engine would pull
several websites that could help us answer this question. On the
other hand, generative AIs such as ChatGPT will pull up the
most likely answer to respond to the question.

ChatGPT will also provide lists of seminal authors, studies, and

experiments done within a certain field. This is in contrast to
the search engine which will pull up websites or documents
with lists that others have made. The search engine will not
consolidate results. In addition to getting one consolidation of

23
results, ChatGPT will also provide a one-sentence summary of
what it pulled as a result.

ChatGPT’s response may have a perceived authority that could

be harmful. Search engines used to offer multiple sources for a
response, which sometimes revealed other sources, people,
experiments, or work that were not in most lists. This often
unintentionally broadened our original thinking beyond our
initial search. Since ChatGPT gives answers and not sources to
explore, it often does not automatically point us in new
directions without being prompted to do so.

Sample ChatGPT prompts:

1. “Jean Piaget developed the theory of cognitive

development. What are some seminal scholars, studies,
articles, and books that address this field?” / “How did
Lev Vygotsky add to Piaget’s work?” / “Why do you
think Vygotsky focused on the sociocultural impact on
cognitive development and Piaget did not?” / “Describe
contemporary perspectives on cognitive development
and how they connect with Piaget and Vygotsky’s work.
Who is conducting this research now?”
2. “How did Silent Spring by Rachel Carson influence
today’s environmental movement?” / “Provide an
example of another seminal book or study in
environmental science and activism that was impacted
by Carson’s work.” / “How do current environmentalists
view the work done by Carson and Ehrlich?” / “Provide
a list of studies that further explore this topic.”

24
 13
Evaluate: Correctness
Assign students to generate a specific output, then
explain whether ChatGPT’s output is factually
correct.

Because ChatGPT is a large language model of generative AI, it

relies on language prediction, rather than analysis of facts, to
create sentences. In other words, it generates each next word as
the “most likely” to fit into the sentence thus far, based on all
the other language it’s ingested in the past. As a result, the
system doesn’t actually think and certainly doesn’t understand
what it’s producing. As defined in the Introduction, LLMs are
parrots of sorts in that they mimic speech without truly
understanding it. Because each word is essentially a guess,
wrong words can easily sneak into any sentence. The system can
generate details and facts that are essentially invented, and even
present conclusions that are flat-out wrong. Students need to be
educated on the weaknesses of the software and trained not to
rely on it unquestioningly.

To many students, ChatGPT’s output looks convincing. Indeed,

the system is designed to provide answers with absolute
authority and, as a result, has every appearance of being true. Its
verisimilitude (or, as some might say, its “truthiness”) is a
deception, however. Students are better advised to regard
ChatGPT output with suspicion, as if potentially every line was

25
a lie—almost as though the software was playing a version of
the game Balderdash, where the point is to create realistically
sounding words and definitions of things that might be
completely invented.

Assignments could be as simple as giving students the

parameters of a ChatGPT prompt, and then asking them to
evaluate how true each part of the output is. Such a task is made
even richer if students are required to provide external
corroboration of the correct answer, such as from other assigned
readings or videos. One variation might be where the student
asks ChatGPT to role play as an interviewee and rates the
correctness of the responses.

Sample ChatGPT prompts:

1. “You are now a nuclear physicist at a press conference.

The first question is how magnetic resonance imaging
functions.”
2. “Cite the Florida statute that prohibits the harassment of
manatees.”
3. “What are your legal obligations as a social worker if a
client discloses elder abuse to you?”
4. “Is it a violation of FERPA to discuss grades with a
parent over the phone?”

26
 14
Evaluate: Hallucinated Sources
Require students to verify the existence of quoted
sources via the institution’s library website.

ChatGPT will readily provide sources, including academic or

refereed ones, upon request, and can also provide a bibliography
in any citation style requested, such as APA, MLA, or Chicago.
However, as an LLM, it is vulnerable to “hallucinating”
(inventing) facts in general, which can include sources,
citations, and direct quotations. It’s built to string plausible
words together, not to isolate real quotes. One common result of
a request for citations or sources is the acknowledgment of a
real scholar, perhaps even the exact ideal scholar an expert
might quote, but with publication titles or journals listed that
sound realistic, yet do not exist. Sometimes real publications are
listed, but with some incorrect details, such as missing authors
or the wrong year of publication.

As part of their broader education in AI fluency, students need

to be reminded that AI-generated sources might not exist. One
idea to accomplish this is to assign students the task of obtaining
an AI-generated bibliography for a targeted essay topic, then
have them upload a screenshot from the website of your
institution’s library database where they have verified whether
the listed source exists or not, repeating as necessary for all the
sources in the bibliography. A related assignment could
challenge students to dive deeper still by evaluating the

27
scholar’s credibility and activity in this field, and, if valid, then
summarizing (perhaps with ChatGPT’s help) a relevant
publication by that scholar, or possibly even their entire related
body of work. Or students could be asked to generate a
bibliography of real sources—after generating the AI-created
one—as a way to underscore the skills of library-based research
and proper style formatting.

In terms of evaluating ChatGPT-provided quotations, students

could be asked to paste the AI-provided quotation into a search
engine (using quotation marks at the beginning and end to
guarantee an exact match), in order to see whether the quote is
genuine. Additionally, students might also be asked to locate a
real quotation to replace the hallucinated one provided by
ChatGPT.

Sample ChatGPT prompts:

1. “Write a bibliography about the rate of rainforest

depletion.”
2. “Create a bibliography of the five most important
publications on rainforest depletion. Write it in APA
style.”
3. “Summarize the work of W.F. Laurence on rainforest
depletion.”
4. “Provide the top five quotations by academics about
rainforest depletion.”

28
 15
Evaluate:
Soundness of the Argument
Create a student task to evaluate ChatGPT’s
argument for its flow and logic.

To be considered well-formulated, an argument must include

both elements that are true and connective logic that is valid.
While individual facts generated by ChatGPT may be untrue,
this assignment instead assumes that the facts are true but
requires students to examine how those facts are assembled into
an argument. Does the argument’s structure contain any flaws
or overly hasty assumptions?

Examples could include asking students to use ChatGPT to

generate a syllogism (two premises and a conclusion) and then
evaluate whether the conclusion correctly follows these
premises. Or, students might evaluate whether the evidence
presented in an argument is both necessary and sufficient to
prove the argument correct. One additional idea might be to
instruct students to evaluate an argument’s warrant (the usually
unstated assumption about WHY a given statement is assumed
to be true).

29
Sample ChatGPT prompts:

1. “Present a 250-word argument that no-fault insurance

laws unfairly impact lower-income drivers.”
2. “Write a three-paragraph persuasive essay that science
argues for the existence of God.”
3. “Create a syllogism that claims turkeys would have been
a better national bird for the United States than the bald
eagle.”

30
 16
Evaluate: ChatGPT’s
Predictions and Extrapolations
Instruct students to generate a ChatGPT
prediction and then assess the output.

One of ChatGPT’s strengths is its ability to extrapolate, even in

hypothetical situations, and present what appears to be realistic
prognostications. Yet all the caveats about ChatGPT’s output
with regard to factual accuracy and logical argumentation
continue to apply, since its method of generating output is just
as suspect when creating predictions. Students should be trained
to regard these predictions with the same level of guardedness
as they do the output by ChatGPT which is supposedly more
factual. However, because predictions cannot be fact-checked
with a simple Web search, a more analytical approach is called
for, where students apply reason and previous learning to the
problem as they attempt to diagnose the prediction.

The process begins with students being asked to generate a

prediction via ChatGPT, ideally one the instructor has specified
in detail and has previously vetted. The students should then be
instructed to apply both previous knowledge and critical
thinking to evaluate the output. Possible areas of evaluation
include completeness, likelihood, and veracity.

31
Sample ChatGPT prompts:

1. “Imagine a world where every human baby was

implanted with a microchip to enhance cognition and
recall. How would this affect the job market?"
2. “What would be the human toll on worldwide ocean
levels rising by five meters?”
3. “What will be the impact on Americans with low
incomes if Social Security can no longer pay full benefits
in the future?”

32
 17
Evaluate: Logical Fallacies
Direct students to create an argument in
ChatGPT, then identify any logical fallacies.

Since ChatGPT isn’t creating arguments built around logic (or

even rules), it’s possible for the software to create arguments
that replicate logical fallacies in text it’s been previously trained
on. As the expression goes, “garbage in, garbage out.” The
system is only as good, or as foolproof in logic, as the material it
was trained on. Since ChatGPT was trained on a very wide
variety of inputs, some of the material may have contained
logical fallacies from the start.

If the course material already includes learning about logical

fallacies, using a large language model in an assignment makes
perfect sense. Students could be asked to scan for the presence
of straw man, post hoc, slippery slope, hasty generalization, or
false dichotomy logical fallacies in the AI-generated output,
ideally with a detailed explanation of how the fallacy occurred,
and what could be done to prevent it.

33
Sample ChatGPT prompts:

1. “What are the most common arguments for and against

the use of fracking?”
2. “What are the arguments for and against gender-neutral
bathrooms in high schools?”
3. “Some people claim the singularity is near, promising
immortal human consciousness. What are their central
claims?”

34
 18
Evaluate: Write a Rebuttal
Tell students to first obtain an argument from
ChatGPT, then craft a rebuttal.

Since ChatGPT creates its arguments based on the material it’s

been trained on, it very much constructs the elements of its
arguments on ideas previously laid down by humans. As such,
its logic is neither impervious nor sacrosanct. Incomplete,
incoherent, or even patently false claims and conclusions can
become part of ChatGPT’s output, though sometimes they are
disguised as lesser points or optional perspectives.

Consider instructing students to upload the complete prompt

and output of their query to ChatGPT, followed by a detailed
rebuttal. A suggested best practice is to scaffold the rebuttal for
students into discrete parts, so they are not confronted with a
task too large initially. They could be asked, for instance, to first
address the scope of the output—does it leave out critical areas,
or focus inappropriately on others? Then, they could be asked to
weigh in on the conclusions rendered. Do they seem
appropriate, or out of scale? Would alternate conclusions have
been better? What evidence is missing from the argument
presented?

35
Sample ChatGPT prompts:

1. “Explain in 3 to 4 paragraphs why paying your own way

on a first date is the best choice.”
2. “If a modern-day time traveler were to journey to the
Middle Ages, what should they do to improve the
quality of life around them since they have the benefit
of future knowledge?”
3. “Advance the argument that although electric vehicles
do not consume gasoline, their carbon footprint is still
significant, even when compared to gasoline-powered
cars.”

36
 19
Evaluate: Bias
Assign students to prompt ChatGPT to provide a
detailed output (ideally on a topic you pre-select),
then require them to evaluate to what extent the
software’s answer might contain bias.

As a product of being trained on human-generated content,

databases, and raw data, ChatGPT will naturally reproduce the
biases, assumptions, or stereotypes that humans who created
that content held, or otherwise appear in the data. Examples
might include systemic racism or unconscious bias. Recognizing
the potential for harm in simply reproducing bias, ChatGPT’s
creators continually make tweaks to the algorithm, and some of
its “natural” answers are prevented. This laudable effort can,
however, have perverse unintended consequences when
mitigation efforts are applied partially rather than universally.
For instance, in the same session, ChatGPT once refused to “tell
me a joke about women,” but offered no objection when
prompted to “tell me a joke about men.” This mixed and
complicated backdrop means AI output could contain biases of
various types. Knowing how to assess and evaluate AI content
for bias is a core component of AI fluency. Indeed, the ability to
ferret out bias in AI-generated content is likely to be a seminal
workplace skill for decades to come.

37
ChatGPT can alternately be used to address the inverse
problem; namely, that bias present in human-created text could
be found by using the software. This can be accomplished by
inputting one’s own writing in the system along with a prompt
to identify any biases or ways the content could rub readers the
wrong way, even if unintentional.

Sample ChatGPT prompts:

1. “What is the history of critical race theory?”

2. “Who discovered the DNA double helix?”
3. “Give an example of a model minority.”
4. “Analyze my writing for bias or poor optics—anything
that could be received the wrong way.”

38
 20
Analyze: Summarize
Longer Texts
Encourage students to use ChatGPT to create
summaries of longer texts.

While it may seem counterintuitive to the aims of higher

education to encourage students to seek summaries rather than
read longer primary texts, the general drift in recent years and
recent generations has been away from reading longer texts, and
the arrival of generative AI likely signals the final switch away
from longer readings and toward AI-powered summaries. As
such, it seems wise to embrace the change as it begins, rather
than asking students to cling to traditional methods of education
even as the tools around them change dramatically.

This task represents asking ChatGPT (rather than the students)

to provide analysis. Though this phrase may sound innocuous, it
represents a major shift in education-related tools. Until the
advent of generative AI, it was customary for tools to provide
shortcuts in mathematics or even linguistic expression, but the
task of analysis was largely left to the student directly. With
ChatGPT, it becomes possible and likely even desirable to leave
the initial analysis to the AI. Among other things, such a shift
heralds the beginning of a new era in efficiency. With human
agency not needed as immediately in the initial steps of analysis,
this frees people to focus more on executive function, higher-

39
order thinking, and deeper levels of analysis than ever before,
all while achieving the same level of analysis faster than ever
before. Students need training on this kind of AI analysis, as it’s
still very new in their educational experience.

The permanent addition of artificial intelligence to our set of

tools for accessing information and sense-making very likely
changes the rules. Strategies for active reading such as the SQ3R
method (survey, question, read, recite, review) might well look
different when students can ask ChatGPT to summarize longer
texts for them. While educators want to be careful not to cede
all information-holding to external sources (we want our future
engineers and doctors to have memorized critical things, after
all!), we might need to change the quantities and proportions of
what does get memorized. This sort of shift is always underway
in society. Most modern drivers of cars would not know how to
fix a roadside breakdown, but such skills were more common
decades ago when motors were simpler and more mechanical.

Sample ChatGPT prompts:

1. “Isolate the top 3-4 ideas from this article and list them
as bullets.” [Paste text below].
2. “Provide a summary of the following story/chapter.”
[Paste text below].
3. “tl;dr of the Magna Carta” (tl;dr is slang for “too long;
didn’t read,” which is a shorthand request for a
summary)
4. “Create ten possible headlines for social media posts that
will entice readers to click to view my article.”

40
 21
Analyze: Find the Needle
in the Haystack
Show students how to ask for a specific piece of
information in a much longer text.

One of the most promising features of AI is its ability to

automate tasks, and in the process save a lot of time. Students
who know that a particular piece of information is explained
somewhere in a larger text, but do not know how to find it,
could save a great deal of time by pasting the full text to
ChatGPT and asking ChatGPT to locate, quote, or otherwise
summarize the needed information.

There is a pedagogical consideration here that may complicate

the utility of AI as an advanced search agent. When students
engage a lengthy text with their eyes and scan it for
information, there are several important acts of cognition taking
place—and many of them are responsible for forging
connections, making meaning in context, and even forming
long-term memories that can later be retrieved. When reading
becomes more passive, the details are likely to be less retrievable
from long-term memory. Educators should consider carefully
whether summarization—or information shortcuts—fit their
pedagogical aims for that assignment.

41
Sample ChatGPT prompts:

1. “Who has the right to vote, according to this text?”

[Paste text below].
2. “Who kills Rosenkranz and Guildenstern?”
3. “What does the Declaration of the Rights of Woman say
about property?”

42
 22
Analyze: Critique and Interpret
Encourage students to use ChatGPT to provide an
interpretation of a text.

One of the primary skills desired by many disciplines is critical

thinking, especially in interaction with a text. For students
without extensive training in text-based critical thinking, the
desired thought process can be elusive. ChatGPT offers a way
for students to quickly see examples of interpretation, which
may provide models they can more readily emulate. As with
summarization, however, there are pedagogical dangers here,
especially if students become over-reliant on the software.
Instructors may want to include assignments that replicate these
learning outcomes without the use of ChatGPT, after the AI-
fueled scaffolding is accomplished. To prevent the use of
ChatGPT on these additional assignments, they might be best
accomplished in an in-person setting.

Students might be asked to examine a text for its persuasive

argument. When engaging ChatGPT for assistance, they could
consider seeking an analysis of the structure, soundness,
plausibility, or originality of the argument. These same skills
could be employed for literary texts and figurative speech like
poetry. In those cases, ChatGPT could be used to help isolate
important themes and messages of the work.

43
Sample ChatGPT prompts:

1. “What does this poem mean?” [Paste text below].

2. “Identify the major themes of this chapter.” [Paste text
below].
3. “What are four common interpretations of Camus’s
Stranger?”

44
 23
Analyze: Poetry and Figurative
Language
Allow students to use ChatGPT to begin the
process of interpreting a poem’s meaning while
building their understanding of figurative
language and its uses.

According to the National Endowment of the Arts’ Survey of

Public Participation in the Arts (SPPA), only 12 percent of U.S.
adults read poetry (or listened to it through recordings,
broadcasts, or web streaming) in 2022. And while Instapoets
have boosted the popularity of poetry, some book retailers
estimate poetry collections still only account for one percent of
book sales.

There are countless reasons why many dislike poetry, including

negative school experiences where poetry was dissected and a
student’s understanding of its meaning was assessed, a general
preference for literal texts, and a dislike for work that’s
considered pretentious. But the process of analyzing a poem
(and other mediums that use figurative language) can lead to
many benefits, including improved literacy and reading
comprehension, higher levels of emotional intelligence, and
stronger critical thinking skills.

45
ChatGPT can summarize the literal meaning of a poem (stanza
by stanza) and offer historical and cultural context that provides
students with comprehension clues on a poem’s theme or
deeper intentions. Students can also ask ChatGPT to evaluate
their own analyses, check for biases or shifts in tone they may
have missed, and provide background on unfamiliar terms or
references.

Sample ChatGPT Prompts:

1. “Summarize the literal meaning of each stanza of this

poem while identifying key poetic devices. [Paste text
below]. Then, highlight repeated words, themes, or
sounds, connect themes to cultural or historical events
or contexts, and find poems with similar themes.”
2. “Analyze my analysis of this poem. [Paste text below].
Do you agree with my interpretation? Are there any
gaps in my argument? Can you play devil’s advocate and
provide an alternative interpretation?”
3. “Create a game that uses figurative language to teach the
concept of intermolecular force fields. Use alliteration,
hyperbole, idioms, metaphor, and simile.”

46
 24
Analyze: Convert to
Conversational Language
Encourage students to use ChatGPT to simplify
language when reading complex texts or learning
difficult concepts.

When learning new, challenging concepts or reading dense

texts, students often struggle with comprehension due to
academic language and discipline-specific styles, especially
when those materials assume prior foundational knowledge
students may lack. For years, students have been turning to
YouTube and other platforms to fill these gaps, but this can also
lead them to inaccurate or misleading information.

Using ChatGPT to convert these readings or concepts into text

that uses conversational or “lay” language can act as an
accessible starting point for more thorough comprehension. It
can also reinforce learning, validate their own understanding,
and make concepts relatable to their own experiences.

Students can also use ChatGPT to practice their ability to

explain their own research or projects in ways that those outside
their discipline can easily understand. By inputting their work
into ChatGPT, students can check the effectiveness of their
communication and writing skills. They can also ask ChatGPT

47
to list the techniques it used to simplify the text, furthering
their ability to rhetorically analyze the conversion while
learning more about their discipline’s conventions (i.e., jargon,
sentence length, passive or active language, etc.).

Sample ChatGPT prompts:

1. “Convert this text on the differences between

differential and integration in calculus in a way that
someone who has not studied this level of math can
easily understand.” [Paste text below]. / “Can you
provide examples of how these concepts show up in
everyday life?” / “Provide a more thorough description
of how calculus is used for GPS navigation.”
2. “Analyze this text I wrote about the hippocampus and
neurogenesis by creating bullets of key points.” [Paste
text below]. / “Convert those bullets into a few
paragraphs that a layperson could easily understand and
provide examples of how this could impact a person
directly.” / “List the techniques you used to convert the
text to be more understandable to a layperson.”

48
 25
Analyze: Qualitative Analysis
Require students to use ChatGPT to perform
qualitative analysis, then evaluate the output.

ChatGPT excels at identifying and summarizing patterns of

words, phrases, and even conceptually related words that
otherwise seem disparate. As such, it performs qualitative
analysis quite well. It can be asked to identify repetitions,
themes, and broad categories, as well as identify individual
words or topics that stand out. This is true even when the
writing being analyzed is careful with its diction to avoid
repetition, because ChatGPT can identify synonyms and group
them into logical categories. Specifically, it can apply topic
modeling techniques like Latent Dirichlet Allocation to discover
latent topics within a text, or extract names of people,
organizations, locations, etc.

In addition to categorizing texts and understanding prevalent

topics within the content, ChatGPT can work more holistically
to provide analysis of the overall sentiment of a piece of text or
identify opinions present in the writing. Students could be
asked to provide their own judgment about ChatGPT’s analysis,
weighing the output on the basis of accuracy and veracity.

49
Sample ChatGPT prompts:

1. “Analyze the tone and sentiment of this letter.” [Paste

text below].
2. “Create a semantic analysis of this article and generate
an editable comparison table.” [Paste text below].
3. “Inspect the phrasing of this text and identify
conceptually related words that repeat.” [Paste text
below].

50
 26
Analyze: Quantitative Analysis
Direct students to perform certain quantitative
analyses at ChatGPT.

While ChatGPT is currently limited to processing only text-

based input, and cannot directly access databases or perform
complex computations, it can assist with basic statistical
analyses such as mean, median, mode, standard deviation, and
variance. Additionally, it can perform simple calculations like
arithmetic, percentages, currency conversions, and time
calculations.

However, some caution is called for here. ChatGPT itself

cautions that its output could include inaccuracies or errors,
which in practice include mathematical errors. In fact, its ability
to calculate might be better used as a vetting exercise, where
students double-check ChatGPT’s math. If it does make
mistakes, students demonstrate their mathematical
comprehension by finding it, explaining why it’s wrong, and
correcting it.

Some calculations can be performed in multiple ways; students

might also be asked to suggest variations to the calculations that
might also work, or even work better. These types of exercises
enforce student learning without relying on automated help or
risking students automatically accepting hallucinated
calculations.

51
Sample ChatGPT prompts:

1. “Provide the mean, mode, and standard deviation for

this dataset.” [Paste text below].
2. “What percent of 178 is 12?”
3. “What is the correct order of operations, and the correct
answer, for 5 + 8 * 6 - 7 / 4?”

52
 27
Writing: Brainstorming a Topic
Encourage students to brainstorm ideas about
their essay topic using ChatGPT.

One of ChatGPT’s greatest strengths is its ability to create lists

and bullets related to almost any topic. This is particularly
useful when asking ChatGPT to break down a complex topic
with many moving parts into constituent fragments. These sub-
components of the main topic can be useful for generating a
thesis or conceptualizing the flow of the argument across
paragraphs.

In many cases, the list will function as a double-check rather

than providing direct inspiration for all elements, but it is not a
rare occurrence for one of the brainstormed ideas to be
something the writer had not thought of previously. One trick is
to also ask ChatGPT to create a list of phrases, which could
cause different creative sparks in the writer.

53
Sample ChatGPT prompts:

1. “List words and phrases associated with the da Vinci

surgical system.”
2. “How do I employ Edward de Bono’s ‘six thinking hats’
about Amazon basin deforestation?”
3. “What are the major factors contributing to the rise of
mental health issues in college students?”

54
 28
Writing: Overcoming
Writer’s Block
Tell students to draft an opening paragraph with
ChatGPT to avoid the paralysis that accompanies a
blank page.

Because writer’s block can be crippling, students are sometimes

told to start writing informally, even freewriting that is more
stream-of-consciousness than complex prose, because additional
writing becomes easier after they’ve eased into a writing
mindset. However, ChatGPT can provide a different starting
point. If the software creates an early draft, students can focus
their efforts on revising rather than composing, which is
generally an easier onramp to the writing mindset than drafting
from scratch would be.

However, students need to be instructed very clearly that such

AI-generated assistance should serve only to inspire and
properly calibrate the writing mindset. The output itself is not
to be used, as this represents plagiarism. Even revising the
sentence with all new words would likely still qualify as
plagiarism, since plagiarism includes stealing ideas and the
ordering of arguments, not just the ordering of words. It’s a fine
line. ChatGPT can suggest ideas and even bullet points of entire

55
arguments, but the final argument itself must flow from the
student’s pen.

Sample ChatGPT prompts:

1. “Write the opening paragraph of a persuasive essay

about the disadvantages of no-fault car insurance.”
2. “Compose the conclusion to a lab report about the redox
reaction of iron and copper sulfate.”
3. “Provide the first half page of a short story about
discovering a lost pyramid in the jungle.”

56
 29
Writing: Request Definitions,
Synonyms, and Antonyms
Instruct students to use ChatGPT as a thesaurus.

Due to its nature as a large language model, ChatGPT is

excellent at manipulating words. Even though it assembles
sentences by guessing rather than truly understanding the
meaning of words, the final result is always fluent and usually
completely comprehensible. It’s just as impressive at the
individual word level. Asking ChatGPT to define a term will
deliver a satisfying output that is rich enough without needing
to be re-prompted for a longer answer (though of course, it will
comply if asked to expand). The same is true for generating
synonyms and antonyms.

Asking students to generate definitions, synonyms, and

antonyms doesn’t initially seem like a worthwhile assignment to
give to students, unless your class happens to focus on words or
linguistics. Primarily, this ChatGPT skill becomes useful in the
act of writing, such as composing an essay. At times during the
composition process, writers may face momentary mental
blocks. They sense there is a perfect word to fit the sentence
they are looking for, but cannot seem to summon it on their
own. Treating ChatGPT like a thesaurus could provide the
means to overcome the mental block. Simply informing
students of this option could work, but it might also be

57
worthwhile to assign a small amount of points simply for
performing the task, perhaps as a discussion board post. In this
fashion, students gain firsthand experience of ChatGPT’s
usefulness while composing, making them more likely to use it
on their own. The actual output of the assignment in this case is
less important than imparting the lesson of ChatGPT as a
thesaurus.

Sample ChatGPT prompts:

1. “Define misanthrope.”
2. “Give me five sophisticated synonyms for foul-
smelling.”
3. “What is the opposite of altruistic? Provide six
examples.”

58
 30
Writing: Generate a Thesis
Require students to paste the essay assignment
prompt into ChatGPT and request a thesis.

Creating a thesis with ChatGPT accomplishes several things.

First, students obtain a firsthand glimpse into the system’s
capabilities (and limitations), which will prove useful for their
future searches. Second, students might find helpful ideas for
the paper they are assigned to write, though they should be
cautioned against simply copy-pasting the ChatGPT output into
their essay and representing it, untouched, as their own work.
Third, students should be given the freedom to attempt multiple
prompts to learn what works best. Slight variations on the
topic(s) requested can have outsized effects on the output, as can
including parameters such as “original,” “ambitious,” or “non-
obvious.” In this fashion, students log incremental gains in their
fluency with writing AI prompts.

Another useful activity regarding software-generated theses is

to require students to include the output with their own work
and critique it. This is a form of human-machine peer editing
that provides some of the same benefits as human-human peer
editing; namely, that in analyzing the work of another entity,
each student gains a deeper appreciation about what is (and
what isn’t) working in their own writing. It also provides
needed reflection time for students to consider the strengths and

59
weaknesses of ChatGPT’s output, making them less likely to use
it for inappropriate cheating in other contexts.

Sample ChatGPT prompts:

1. “Generate a thesis for a five-page essay combining a

feminist interpretation and a Utilitarian analysis of
Marvel’s movie Avengers: Infinity War.”
2. “Generate a thesis for a five-page essay combining a
feminist interpretation and a Utilitarian analysis of
Marvel’s movie Avengers: Infinity War. Ensure the
thesis offers a bold and ambitious argument.”
3. “Generate a thesis for a five-page essay combining a
feminist interpretation and a Utilitarian analysis of
Marvel’s movie Avengers: Infinity War. The argument
needs to be bold and ambitious, and also needs to
advance ideas far beyond the obvious.”

60
 31
Writing: Further Develop
the Thesis
Allow students to practice writing thesis
statements by using ChatGPT as a sounding board.

Like the stages of most projects, the initial construction of a

thesis statement will not be the final attempt. As we discussed
previously, creating a thesis statement can range from being a
simple tentative thesis and evolve into one that is more
complex.

After students have critiqued the theses ChatGPT generated, it

may seem tempting to simply offer them instructions to “Use
ChatGPT to create a variety of additional thesis statements,” but
those instructions lack guidance and clarity. Provide students
with specific, pointed statements and probing questions to use
with ChatGPT. For example, tell students to instruct ChatGPT
to create a thesis statement that expresses the point they want to
make by making a promise to an audience. Another example is
to tell ChatGPT to create a thesis statement in two parts where
the first states the topic, and the second makes a claim about the
topic. An additional option is to have students write what they
think is a good thesis; submit it to ChatGPT, and ask questions
such as, “Will this thesis statement engage my target audience?
Is this thesis statement debatable? Is this statement a fact or
claim?” Your students will be pleasantly surprised when

61
ChatGPT informs them that their statements are either good
thesis statements or that they are not thesis statements at all. If
they are, indeed, not true thesis statements, encourage the
student to ask ChatGPT to help them develop their attempted
thesis statements into actual thesis statements. It may take them
a couple of tries to practice getting it right. Kudos to them. They
are learning!

Sample ChatGPT prompts:

1. “Create a thesis statement that makes a promise to an

audience by listing reasons I will limit social media use.”
/ “Use an ‘I will’ statement.” / Don’t list the reasons.
2. “Is this thesis statement debatable? ‘Grass is green.’” /
“Help me develop this into a thesis statement.” / “Help
me develop this into a one-sentence thesis statement.”
3. “In one sentence, create a two-part thesis statement
where the first part of the statement states the topic
‘limiting social media use’, and the second part of the
thesis statement makes a three-point claim about social
media use.”

62
 32
Writing: Generate an Outline
Require students to use ChatGPT to scaffold their
ability to create and develop outlines for their
writing.

A baker would not bake a pie, dulce de leche, or bahn bao chi
without a list of ingredients and a recipe. A complete list of
ingredients and precise measurements are imperative in baking.
The same holds true for writing. The elements of an outline are
akin to a list of ingredients and precise measurements. Outlining
is one step in the writing process that should not be skipped, but
students often miss the value in this step. ChatGPT offers a
noteworthy outlining companion. While the outline ChatGPT
produces may seem a bit generic, it is actually a great starting
point for students who are still learning how to write.

Students can use ChatGPT to assist them in organizing their

information and seeing connections with their ideas. Let’s say
students have chosen four key points they think relate to their
topic. They can now use ChatGPT as a companion to aid them
in outlining those four key points. For example, students can tell
ChatGPT to list the key points in an order that makes sense.
ChatGPT will offer them a skeleton that now needs a bit more
substance. ChatGPT will likely define all four elements for the
student and then provide a generic definition of each key point.
In doing this exercise, students have used a topic to brainstorm,

63
define, and outline key points. They have even acquired
definitions and background information. Require students to
submit ChatGPT’s output as part of their brainstorming/notes. If
students already have their key points organized, they can
simply ask ChatGPT to outline their key points. Their final step
should be to enhance the outline—using ChatGPT as a starting
point—by adding supporting details under each key point.
Instruct them not to use ChatGPT for this final step.

Sample ChatGPT prompts:

1. “Tell me how wind mitigation, electrical, plumbing, and

HVAC relate to real estate.” / “Now create an outline
about wind mitigation, electrical, plumbing, and
HVAC.” / “What elements are important for a home
inspection in a real estate transaction?”
2. “Create a problem-solving outline for a mathematical
equation.” / “Create a mnemonic device to help me
remember this.” / “Outline the steps I would need to
complete problem X.”
3. “Some of the major disadvantages of a startup company
include difficulty in accessing the market, team
composition, limited resources, and inefficient
processes. Create an outline using these key points.”

64
 33
Writing: Create
Counterarguments
Introduce counterarguments by engaging in a
debate with ChatGPT.

It goes without saying that ChatGPT can be used in a variety of

ways. As it relates to writing, perhaps, one of the hidden gems
of using ChatGPT is in forming a partnership with it. Used as a
co-pilot, it can prove to be an adroit debater. With this
knowledge in hand, students can use ChatGPT to learn how to
familiarize themselves with counterarguments while developing
their own perspectives. As a writer, it is paramount to
remember that other perspectives exist, and they actually
matter. It is important to teach students to recognize and seek
counterarguments early in their writing journey. While writers
may not necessarily agree with the counterargument of another,
it is still important to acknowledge other arguments. By
considering others’ positions on an argument, writers offer more
validity to their own position by showing they have done their
due diligence in listening before joining a conversation.

What better way to join a conversation than to have prior

knowledge that demonstrates the student’s understanding of the
current discussion surrounding the topic and the legitimacy of
the student’s own well-rounded and well-thought-out position.
For example, assign students debatable topics in the form of

65
advertisements, well-known cinematic debates, or even rights
on college campuses. Argumentative topics can be input into
ChatGPT with specific intentionality or as general posed
questions.

Sample ChatGPT prompts:

1. “Argue against the 1946 Camel cigarettes ad that

implies, ‘More doctors smoke Camels than any other
cigarette.’” / “Present an argument in support of the
1946 Camel cigarettes ad that implies, ‘More doctors
smoke Camels than any other cigarette.’”
2. “Which offers a better movie-going experience: Marvel
Studios or the DC Extended Universe (DCEU)?” /
“Present an argument on the position of Marvel Movies
being lighter and more interconnected than DC
movies.” / “Are DC movies darker and more standalone
than Marvel movies?”
3. “Should free speech be protected on college
campuses?” / “Should college professors be afforded
intellectual freedom in their research and teaching?”

66
 34
Writing: First Draft
Encourage students to use previously practiced
ChatGPT writing steps to develop their rough
draft.

Writing a first draft can be daunting for students, as many of

them do not believe they like writing to begin with. If they’ve
made it to this stage, they are likely exhausted if the instructor
hasn’t used the previous helpful tips this book offers for using
ChatGPT to teach the writing process. Encourage students to
create an outline in ChatGPT first (see Tip 32 for instructions).
This outline has already prepared them to write their first draft.
Inform students they are writing an essay of your preferred
length or word count requirement. Remind them of the key
elements needed for this first draft: an introduction, body, and
conclusion, and instruct them on the way you want them to
construct the essay.

This step is where using ChatGPT offers a bit of amusement.

Begin by instructing your students to create an attention-
grabbing introductory sentence (hook) and ask ChatGPT if it
thinks the sentence will truly reel in a reader. They can even
ask ChatGPT to enhance their sentence. Is this any different
than using the provided synonyms option in a Word document
or asking a friend to proofread work?

67
Next, some of the key terms and definitions they were provided
by ChatGPT in their outlining stage can now be used in their
introduction. You probably already see where this is going.
They can even use the thesis statement they created in
ChatGPT! You just scored a big win for the team by using
ChatGPT to assist them in all those steps. Next, they will
develop the body of their essay using whichever mode of
writing you’ve required, their outline, and ChatGPT. Finally,
encourage students to pull their topic sentences together and
write a hearty conclusion using, you guessed it, ChatGPT.

Sample ChatGPT prompts:

1. “Do you think this is a catchy introductory sentence? ‘It

was the best of times; it was the worst of times.’” / “Give
me a better one then.”
2. “How can I use these key terms to enhance my
introduction?”
3. “Now that I’ve written my introduction, help me
develop this paragraph into a more refined chain of
thoughts about Key Point A.”
4. “Here are my topic sentences. Help me pull them
together to create a captivating conclusion.”

68
 35
Writing: Improve Topic
Sentences
Demonstrate how ChatGPT can assist in elevating
your topic sentences by signaling what’s ahead.

Accessible content ahead! This is exactly what topic sentences

offer. Like style and section headings and alt text for images,
topic sentences guide the reader to understand what each
section of a piece of writing is about. They also offer ease in
locating various sections within a written work. This makes
writing so much more accessible and easier to follow. Often,
students forget to add topic sentences to their writing, but
ChatGPT can assist with this. Students can take the key points
from their thesis statement and enter those into ChatGPT. A
series of commands could be used to prompt ChatGPT to
generate refined topic sentences.

It is likely that students either have topic sentences that still

need a bit of work, or they do not have any at all. A couple of
scenarios using ChatGPT could easily resolve this issue. For
example, encourage students to use the topic sentences
ChatGPT generated for them using their key points and add
those to their essays. You could also have students input their
entire thesis statement into ChatGPT and ask it to create a
specific number of topic sentences from the thesis statement.
Have students then enter their topic sentences into ChatGPT

69
one by one and ask ChatGPT to rephrase them. They can even
ask ChatGPT to reword the sentence in two or three sentences
and ask for regenerated options if they are not satisfied with the
ones ChatGPT produces. This is also an opportunity for you to
encourage students to enhance whatever ChatGPT produces
using their own creativity and knowledge. These combined
sentences offer enough information for students to write a
hearty conclusion.

Sample ChatGPT prompts:

1. “Create topic sentences using my 3 key points.” [Paste

text below with individualized key points].
2. “Make my topic sentences better.” / “Give me one more
topic sentence about _____.”
3. “Here is my thesis statement. Help me create four topic
sentences from this thesis statement.” [Paste text below
with specific thesis statement].

70
 36
Writing: Improve Transitions
Create a game using ChatGPT that teaches
students how to mark logical relationships
between ideas.

The purpose of this chapter is to force you to step away from

teaching and learning mode and transition (pun intended) into a
gaming and learning mode. Your students will still learn, and
you will still technically teach, but the fun is in the preparation
and the students reaping the benefits of your game.

As most instructors do not know how to create a video game

using coding and advanced technology, there is a companion
that can assist you in creating something almost equally as fun.
Use ChatGPT to help you create a game or several games that
teach students how to effectively use transitions. These games
could be developed for a mobile app or a web-based platform.
Not only will ChatGPT create an exciting game for you, but it
will also offer valuable learning objectives that likely align with
your curriculum. ChatGPT might also suggest a target audience,
so you can tweak your prompt to align with your level of
teaching. One game ChatGPT created was called Transition
Trek: Journey of Logical Relationships. This title was derived
from the first prompt below, and the game is beyond cool if
you’re into “embarking on a virtual journey through various
landscapes, encountering writing challenges along the way.”

71
You could easily adjust the target audience through the wording
of the prompt or use prompt chaining to solicit a more refined
response from ChatGPT. After using prompt chaining, ChatGPT
created a new game titled, Transitions Mastermind: Unlocking
Coherence in Writing and an additional game titled, The
Transition Carnival, that embraced a fun and lighthearted vibe.
All three games included an overview, gameplay mechanics,
educational objectives, and target audience information.

A final option is to assign your students the task of creating a

game individually or as a group. By way of a vote or randomly
drawn numbers, students’ created games could be selected to be
played in class. Engage with us online by sharing your games
and target audience.

Sample ChatGPT prompt:

1. “Create a game that teaches students how to use

transitions in writing to mark logical relationships
between ideas.” / “For a college audience” / “Make it
humorous” / “Make it a mystery.”
2. “Create a game set in a land populated by students who
speak in such a manner that they abruptly change the
subject and everyone remains confused. Have the game
teach students how to use transitions to connect their
ideas so their conversations make sense.”

72
 37
Writing: Improve Connections
between Claims and Evidence
Create an assignment where students must state a
claim, provide support, and ask ChatGPT to
analyze the validity of their claim with respect to
how universal their assumptions are.

Undoubtedly one of the most important aspects of writing is

offering an explicit claim and supporting it with sufficient
evidence. The claim is clearly stated via a thesis statement and
the evidence is laid out in the body of the essay. Often, the
connection between claim and evidence relies upon
assumptions, and this is where there may be a disconnect
between writer and reader. If the writer has one assumption but
the reader has another, the text won’t come across as logical to
the reader. Since assumptions are often blind spots for authors
who might assume everyone thinks the way they do, this
connection between claims and evidence represents a potential
weak point in argumentation that ChatGPT can assist with.

Students should prompt ChatGPT to analyze their arguments

for hidden assumptions that may not be universally held. While
this seems obvious in politically charged topics, many fields
outside politics contain political minefields.

73
ChatGPT can easily analyze students’ claims to determine
whether they are in alignment with experts already engaged in
the topic’s dialogue. For example, direct students to enter a
claim into ChatGPT along with their proposed evidence and ask
ChatGPT to analyze it. Reviewing the output by ChatGPT is
also a good activity for leading a discussion on the need to
analyze its assertions for biases and to research its validity.

Sample ChatGPT prompts:

1. “Analyze my argument about climate change for

assumptions that may not be universal.”
2. “Cost of Living Adjustments (COLA) seems to offer
minute incremental increases in salaries and wages that
don’t allow individuals to truly keep up with inflation,
but the government continues to offer them as some sort
of noteworthy contribution. Many citizens are still
living below the poverty level, while the rich seem to
keep getting richer. Analyze this claim for assumptions
that not everyone may share.”

74
 38
Writing: Teach Parts of Speech
through Sentence Diagramming
Utilize ChatGPT as a study aid to assist students in
identifying parts of speech.

One of the best ways to teach students parts of speech is

through visual representations. This activity can be daunting for
the instructor as well as the student. Why not make it
competitive? Use ChatGPT to help you create a variety of
sentences that your students can diagram. Furthermore, you can
use ChatGPT as a way to assign students to create their own
sentences, to offer assistance in producing sentences when you
try your hand at asking it to output whimsical or funny
sentences, to create a game that teaches parts of speech through
sentence diagramming, and to set the parameters of the game.

It goes without saying, that most of this chapter is fairly self-

explanatory, but it is important that you don’t lose the real
meaning behind this lesson. Many English Language Learner
(ELL) or English for Academic Purposes (EAP) classes do teach
parts of speech to English language learners. For some, this will
be the foundational learning activity that assists them in
crossing this bridge. Learning parts of speech is the link for
many of the other things they will learn as it relates to learning
the English language. If they can master this activity, they are
well on their way to identifying sentence parts, recognizing the

75
importance of using measured voice inflection, writing
paragraphs, and even more advanced techniques. For those who
are not ELL or EAP students, parts of speech and sentence
diagrams are foundational elements in their educational journey
as well. Whether you create sentences on paper, whiteboards,
SMARTBoards, or anything else, make it fun; make it
competitive; and offer incentives. Students will remember the
day they diagrammed sentences because they had so much fun.

Sample ChatGPT prompts:

1. “Create a game that teaches students parts of speech via

sentence diagramming. Create it for two teams. Make
the sample sentences funny. Suggest some prizes each
team can win.”/ “Where are my sentences?” / “List the
sentences.” / “Give me more prize options.”
2. “Create a lesson plan for a college class that teaches parts
of speech by diagramming sentences.”
3. “What is the best way to teach parts of speech using
sentence diagramming?”

76
 39
Writing: Perplexity
Encourage students to use ChatGPT as a fill-in for
writer’s block and failure to finish thoughts.

Even the most skilled writer experiences writer’s block. While it

may not last for an extended period for all, writer’s block is
common, nonetheless. “Perplexity” is analogous to what a
ghostwriter can do for an author who experiences writer’s
block. Perplexity is a measure of how well a language model can
predict the next word in a sequence of words and measures how
well a system can understand natural language. As these
chapters were being typed, perplexity reared its “pretty” head
(pun intended) and offered sentence completion and word
completion. Perplexity has been used for quite some time in
texting, emailing, document creation, and numerous other
ways.

Ironically enough, some people say they do not use AI and are
not interested in learning anything about it, but they allow
autocorrect on their cellphones, use F7 when checking Word
documents for errors, and check Waze to avoid traffic on their
way home. There is no reason to fear perplexity. As long as your
language model has a low perplexity score, your reader can
better understand your writing. Think about it like this: to be
perplexed is to be baffled or confused. The less perplexing
something is, the more intelligible it is.

77
At an even more advanced level, ChatGPT offers a solution that
provides intercession for writer’s block. OpenAI’s GPT-3.5 and a
large language model (LLM) is an AI system that uses natural
language processing (NLP) and machine learning to provide a
conversational search engine based on AI chat. It can help
students find an array of topics. This is useful because it is like
Google and OpenAI had a baby. Students can enter their
questions into Perplexity AI’s chatbot-like interface and watch
it generate natural language responses.

Instruct students to offer their incomplete thoughts to ChatGPT

and watch perplexity bring life to their writing. It might prove
valuable to have students show their previous work before they
solicited help from ChatGPT. This could help you gauge how
much is their own work versus ChatGPT’s output.

Sample ChatGPT prompts:

1. “How can perplexity be used to enhance my writer’s

block?”
2. “Help me finish this thought about how this would be
an ideal story to add to Chaucer’s Canterbury Tales.”
3. “I am attempting to write a cookbook, and the other
section I want to add is on the tip of my tongue. Help
me figure out an additional section to add.” / “Make it
specific to soul food cooking.” / “Make it specific to
Mediterranean cooking.” / “Sous vide.”

78
 40
Writing: Burstiness
Utilize ChatGPT to enhance writing techniques.

If your students are submitting work that seems to be lacking in

variability and creativity, and all sentences look like they were
written by a robot, it is likely your students have not worked
out all the kinks of burstiness. Burstiness relates to the
variability of word choice (diction) and sentence structure
(syntax) in a given section of text. In writing we caution
students about varying their syntax and using words that more
closely represent their own voice in their writing. But this kind
of advanced writing takes a lot of practice, so students often
cannot imagine what varied diction looks like without seeing
examples.

Here is where ChatGPT can help. The large language model is

built on rules such as burstiness, as well as a similar sentence-
level element called perplexity (that also measures
sophistication and variance). As a result, ChatGPT can respond
to explicit calls to increase burstiness and perplexity and will
offer more advanced prose as a result.

79
Assign students writing prompts on various topics and tell them
to have ChatGPT respond to the prompt. In its true form and
fashion, ChatGPT will produce essays that exhibit low levels of
burstiness. Students should print out their results and bring
them to class, so they can practice enhancing the essay
according to whichever point of emphasis the lecture focuses on
that day. The goal is for students to learn to write as more
experienced writers do by using sentences that are more
complex and parallel to sentences that vary in sentence
complexity and length.

Sample ChatGPT prompts:

1. “Raise the level of perplexity in this pasted text.”

2. “Using a very high level of burstiness and perplexity,
explain the process of cloning.”
3. “Repeat that explanation, but with a much higher level
of burstiness and perplexity.”

80
 41
Writing: Grammar and
Syntax Check
Treat ChatGPT as a spellchecker and syntax
companion.

Checking for grammatical errors and sentence structure are

integral elements that must be included in teaching students
how to write. Many originality checkers and plagiarism
detectors offer a variety of advanced options that will scan a
student’s work for originality, grammar, sentence structure,
subject/verb agreement, etc. ChatGPT, unsurprisingly, rivals the
best of the best when it comes to proofreading a piece of
writing. It checks for everything with a simple click on the
submit button, and it produces all results in one output. There is
no need for multiple clicks to navigate to different pieces of
information. It also offers numerous options for revising the
style of writing.

Students could easily ask ChatGPT to vary the syntax of their

written work, but how much more fun would it be to have
students compete with ChatGPT? With a little help from their
individual (or group) devices, have students ask ChatGPT to
revise a piece of their written work, a well-known work, or one
of their favorite songs using different styles of writing. You
would need to have this planned out in advance. Maybe start
with the more far-removed options such as other people’s works

81
and transition into having them submit their own work.
Sometimes it is easier to see the errors of another’s ways than
one’s own. Lastly, remember ChatGPT may not offer the same
output as one of the premium originality checkers. It simply
corrects the mistakes for you and completes the job. As a final
assignment, ask your students to compare their rough draft to
ChatGPT’s output and track their own changes.

Sample ChatGPT prompts:

1. “Rewrite Emily Dickinson’s poem, Because I Could Not

Stop for Death, using Shakespearean syntax.” / “Now
make it sound scientific.” / “Rewrite her poem using text
speak.”
2. “Proofread my essay.” / “Proofread my essay and track
changes.” / “Check my essay for grammar, spelling,
punctuation, syntax, and originality.”

82
 42
Writing: Analyze
Readability and Tone
Encourage students to triangle peer review using
ChatGPT to cross-check for clarity and tone.

You have likely heard of peer-to-peer work, but triangle peer

reviewing offers ChatGPT as one member of the triad.
Sometimes students may get stuck on how to accurately peer
review certain essential elements of the peer reviewing process.
Surprisingly, students can read a partner’s bad essay and provide
feedback that states it was the best essay they’ve ever read. How
they reach these conclusions is a mystery to most, but including
ChatGPT in the reviewing process can help tremendously.

Prompt students to ask ChatGPT to check their essays for things

such as readability, clarity, tone, etc. Using direct commands
will produce fairly straightforward results. A little bit of prompt
engineering is involved, though. This is where you, the
educator, must guide your student on how to construct a strong
prompt. Consider this, if a student asks ChatGPT if it can
understand their essay, this is not necessarily assisting in
assessing the readability of the essay. You also want to
discourage them from asking questions like, “Is my essay good?”
ChatGPT does not offer personal opinions. It does, however,
offer factual responses that would benefit the students. By using
more structured prompt engineering, like the first example

83
below, students will likely receive a more unbiased, transparent,
and even beneficial response.

Sample ChatGPT prompts:

1. “Check my essay for readability.” / “Check my essay for

tone.” / “What is the tone of my essay?”
2. “Does this song that I wrote clearly convey my intended
message?” / “Why do you think it’s about that?” / “Tell
me more.”
3. “In what tone do you think this piece of writing is
written?” / “I meant for it to sound different.” / “Help
me write it in a more ____ tone.”

84
 43
Writing: Improve Prose
Require students to use ChatGPT to revise their
final drafts using standard academic English (or
the language you require for your course).

Writing styles vary across genres, and the expected outcomes

for a piece of writing vary along with the styles. First things
first, ensure students know what your expectations are for their
writing at the beginning of the term. A fun way to offer
students examples that meet your expectations is by allowing
ChatGPT to assist in creating some sample essays. For your prep,
ask ChatGPT to create three sample essays—one poorly written,
one good essay, and one excellent essay. Make annotations on
each essay explaining what makes each poor, good, or excellent.
You will be able to use these for your introduction to your
writing expectations for students.

Print out the sample essays or upload them to your LMS

(whatever method works best for you). As a class, ask students
to identify which essay is poor, good, and excellent. Require
them to offer support for their stance. Then, share annotations
that have not already been stated during the discussion. Next,
explain to students that at a later point, you will allow them to
use ChatGPT to improve their writing according to the
academic level requirement for your course. You can encourage
them to offer multiple attempts to ChatGPT to see which sound

85
poor, good, or excellent. You can also simply have them use
ChatGPT for guidance on how they can improve their prose in
general by asking a pointed question.

Sample ChatGPT prompts:

1. “Create a poorly written essay on why poor eating

habits are detrimental to your health.” / “Now a good
one” / “Now an excellent one.”
2. “Read my essay and help me improve it.” / “What else
will improve it?”
3. “How can I improve my prose?”

86
 44
Writing: Holistic Editing
Teach students to edit using a well-rounded
approach with the help of ChatGPT.

Editing in the world of academia usually involves students who

do not generally care to fully invest themselves in ensuring the
work they are reviewing is given the best possible review. They
simply desire to meet the minimal requirement and be finished
with the task. As educators, it would behoove us to offer
students a better and more inviting way to edit work.

Rather than spot-editing, instruct your students to generate

specific prompts that ask ChatGPT to check for specific
elements, but all elements must be checked in order to edit
using a holistic approach.

Sample ChatGPT prompts:

1. “Proofread my writing and check it for grammatical

errors.”
2. “Rephrase my sentences to refine the tone and style.”
3. “Check my work for consistency.”
4. “Please offer suggestions for how I can improve the flow
and structure of my work.”
5. “Vary the syntax of my sentences.” / “Offer synonyms
and alternative word choices.”

87
6. “Identify sentences in my writing that are unclear and
difficult to understand.”
7. “Help me create a catchy title and an engaging
introduction.”
8. “Check my paper for originality and help me with my
APA citations.”
9. “Help me brainstorm ideas for character development,
plot twists, and story arcs.”

88
 45
Generate: Elaboration
Instruct students to elaborate on their work by
asking ChatGPT to use a step-by-step process,
offer pros and cons, or compare and contrast.

Using ChatGPT to generate general responses is common, but

students can receive an added benefit by allowing ChatGPT to
elaborate on a wide range of topics. Teach students how to
incrementally adjust the temperature of ChatGPT, use max
tokens, and finally review and refine on their own using their
own brains. Different approaches can be used for this
elaboration. Students can ask ChatGPT to expound on topics by
offering a step-by-step approach, expand on an idea or decision
by offering pros and cons for that decision, and help them
organize by using compare and contrast scenarios.

Start by offering students a list of topics—that you previously

generated using ChatGPT—and instruct them to choose one
topic. Tell them to ask ChatGPT to elaborate on that topic. Then
teach them to adjust the temperature of ChatGPT with a lower
temperature being more focused elaboration and higher
temperatures demonstrating more creative elaboration. Next,
demonstrate how students can shorten and lengthen the
response output of ChatGPT through max tokens. Instruct them
to enter the same response after each adjustment. They will
likely be pleasantly surprised by the varied outputs. Finally,

89
instruct students to review and refine the work on their own to
their satisfaction and liking.

Sample ChatGPT prompts:

1. “Please elaborate on the causes for macular

degeneration.”/ “Elaborate again.” / Elaborate again.”
2. “Expand on the idea of offering free health care in the
United States by offering pros and cons.” / “What are the
pros and cons of offering free health care in the U.S.” /
“Expand more.”
3. “Which is better, organizing my closet by color-coding
or grouping my clothes by type? Use compare and
contrast to answer.” / “Offer a different scenario via
compare/contrast.”

90
 46
Generate: Role Play
Ask students to create personas and then engage
in interview-based conversations that simulate
real-world scenarios.

Role-playing activities can create a more interactive learning

experience by providing students with the ability to “converse”
with their course content. With ChatGPT’s ability to gather
information from various sources, it can create realistic role-
playing personas based on all sorts of user prompts, including
occupations, communication styles, belief systems, expertise,
motivations, and even empathy.

When students are asked to generate conversations based on a

specific topic, they’re required to evaluate the information
ChatGPT needs to successfully respond. This requires them to
apply content and contextualize it for a specific time period,
purpose, and audience. For example, conversations about
unidentified aerial phenomena (UAPs) sightings are occurring
all over the world. But how scientists, Congress members,
intelligence officials, conspiracy theorists, and UFO bloggers
discuss UAPs varies greatly. Asking students to determine a
topic of conversation and its purpose and audience—while also
creating a persona of “who” is communicating and why they’re
qualified to do so—provides students with the opportunity to
practice scenarios they may experience in their fields.

91
Sample ChatGPT Prompts:

1. “Engage in an ongoing interview with me. You are a

social worker, and I am an elderly man who lives
independently. My doctor has asked you to conduct a
biopsychosocial assessment to evaluate my safety living
alone.”
2. “I am a civil engineer who is proposing a cable-stayed
bridge to span over the river. You are a city planner
who needs to approve this project. Applying the
principles of structural, geotechnical, and transport
engineering—and considering the environmental,
economic, and social impacts—justify my design
decision.”
3. “Assume the persona of Simone de Beauvoir, embodying
her beliefs and experiences, and engage in a
conversation with me, a student in 2023, regarding the
role of feminism today.”

92
 47
Generate: Business Ideas
Empower students to brainstorm new business
ideas for products or services related to their areas
of interest.

According to various surveys, up to 75 percent of Gen Z’ers

claim they ultimately want to become entrepreneurs. As digital
natives who grew up influenced by the internet and social
media, they’ve seen success stories of young entrepreneurs and
they value flexibility and autonomy. Two-thirds of this
generation are also committed to social change and claim a
desire to find—or create—jobs that will positively impact the
world.

When students are able to visualize how they’ll apply the

content they’re learning, they’re more engaged. By empowering
them to combine their creativity and personal interests with
course content, they’ll invest more fully in the material and
connect what they’re learning to potential business
opportunities.

Begin by discussing problems, challenges, and opportunities that

exist within the discipline. Ask students to select a topic,
combine it in ChatGPT with a list of their own skills and
interests, and prompt it to brainstorm ideas for products or
services that fulfill a need while maximizing the student’s

93
attributes. Students can evaluate those ideas, add to them,
determine target audiences, and create final projects that turn
ideas into business plans.

Sample ChatGPT prompts:

1. “I’m a student in nonprofit management with graphic

design skills who wants to start a business that would
serve those with the greatest need in my community.
What are some pressing social issues in [location]?
Which products or services would assist those most
affected by those issues?”
2. “A group of nurses plan to offer health coaching with
assessment, goal setting, nutrition counseling, stress
management, fitness training, and chronic disease
management. Brainstorm a list of 10 possible business
names.”
3. “Generate a list of platformer game ideas targeted to
middle school students. Include animal characters with
superhero skills, a fantasy setting, and a theme of
diversity and inclusion.”

94
 48
Generate: Slogans
Encourage students to further their
entrepreneurship goals by using ChatGPT to
brainstorm ideas for slogans, mottos, and other
catchphrases that build brands.

While only some might remember the origin of Nike’s Just Do It

slogan that began in 1988, almost everyone is now familiar with
the widely successful phrase which has become a cultural icon
and a source of motivation. Creating a slogan that truly
communicates the heart of a brand—while being creative and
memorable—is no easy task, but essential for marketing a
successful business.

Whether you’re teaching a capstone course or tying industry

goals into your content, asking students to use ChatGPT to build
on their entrepreneurship goals by generating key branding
elements can help them crystallize their mission statements and
more clearly visualize their proposed product or service.
Instruct students to create prompts that include specifics about
their proposed businesses and couple it with market research on
the community, population, and industry to assist ChatGPT in
generating ideas. Students can evaluate ideas; test them in class,
discussion posts, or small “focus” groups; and refine them
accordingly.

95
Some students may wish to extend this assignment beyond
slogans and experiment with using text-to-image AI tools to
create graphic materials. By assigning this step, you’ll encourage
students to familiarize themselves with a variety of these types
of programs, including Dall-E, Fotor, Jasper Art, Midjourney,
Nightcafe, and Photosonic. Many of them are free or have free
trials.

Sample ChatGPT Prompts:

1. “Using a hip-hop style, generate a slogan for a Study

Buddy app that matches students based on their learning
styles, goals, and habits.”
2. “Create personas of 5 different kinds of customers for
your proposed product or service and put them together
as a focus group. Test a series of brand-related materials
that you generated and ask for individualized analysis
based on customer demographics and other factors.”

96
 49
Generate: Poetry as Application
Show students how to apply key concepts by
generating content-related poems.

Language arts educators have long used poetry to build literacy

skills in children. With its musicality, phonemic awareness, and
sense of play, young readers engage in a poem’s narrative and
grow their vocabulary while improving their reading fluency.
But many educators also incorporate poetry into other subjects
to encourage creativity, enhance contextual analysis, and boost
critical thinking skills.

Many STEM subjects, for example, explain the natural

phenomena of the physical world by using abstract, complex
concepts. Instructors often begin teaching these concepts by
relating them to metaphors, a key poetic device, that takes a
word, phrase, or concept and applies it to an object or action in
a way that isn’t literally true. One familiar, broad metaphor is
“the greenhouse effect” which describes how the Earth’s
atmosphere traps heat in the same way glass panes in a
greenhouse do.

Asking students to apply poetic devices to concepts they’re

learning can clarify content and enhance engagement. ChatGPT
can be used to create metaphors, provide prompts for poems,
generate poems that seek to distill complex ideas or theories,
and evaluate students’ poems for accuracy.

97
Sample ChatGPT prompts:

1. “Using the voice of Albert Einstein, write a poem that

employs metaphor and sound devices to describe
quantum mechanics. Communicate how he once felt
and how he might feel today, if he returned to Earth.”
2. “Consider the poem ‘Triple Moments of Light and
Industry’ by Brenda Hillman that depicts the oil
refineries in California. Summarize her argument and
generate a list of all the scientific hypotheses, theories,
and laws her poem alludes to.”
3. “Using iambic pentameter, write a persona poem in
which the speaker is the concept of Prescriptivism who
is making an argument against the personification of
Naturalism.”

98
 50
Generate: Social Media Posts
Instruct students to generate audience-specific
social media posts.

According to the Pew Research Center, approximately seven out

of 10 Americans use social media. While young people may
have started this trend, adults of all ages are now connected
through a variety of sites… and advertisers are well aware of
resulting marketing opportunities. In a study conducted in 2023
by Forbes, 77 percent of small businesses—and 97 percent of
Fortune 500 companies—market with social media.

Advertisers aren’t the only ones tapping into these mediums.

Government agencies and nonprofit organizations also use social
media to inform users. When tragedies occur, more than half of
us learn about it first from a social media post.

Instructing students how to evaluate and communicate through

social media can teach them how to better understand concepts
while making that content relevant to specific audiences, in
their field and outside of it. ChatGPT takes content and uses it
to generate targeted social media posts that match the tone and
style of various platforms. Students can evaluate stylistic
differences and analyze how content is shaped and its
effectiveness. When students engage in this process, they learn
how course content is relevant to everyday life and the
importance of shaping it for various audiences.

99
Sample ChatGPT prompts:

1. “Using a paleoanthropology lens while making it

culturally relevant to this time period, write a series of
social media posts for this content: [insert]. Write a post
for Facebook, LinkedIn, Instagram, Twitter, TikTok, and
YouTube.”
2. “Make a list of subjects I could post about on social
media if I wanted people to be more aware of somatic
science in dance and how it’s relevant to everyday life.”
3. “Actuarial science could help governments use data to
better address challenges, like climate change,
healthcare, and agriculture. Write 10 clickbait headlines
to motivate constituents to contact their elected
officials.”

100
 51
Generate: Creative Writing
as Application
Combine creative writing activities with AI tools
to bring students’ ideas to life.

Since the conception of language, people have used the art of

storytelling for educational purposes. Stories activate the
sensory, emotional, and cognitive regions of our brains and
create stronger associations and connections with the content.
Today, numerous pedagogical studies show this practice engages
learners, invites curiosity, and makes content more relevant,
memorable, retrievable, and transferable.

Subjects and courses in the arts and humanities discipline, like

history and literature, may be natural fits for creative writing
activities, but any type of course can use storytelling to engage
and teach students. When students are asked to create their own
stories related to course content, engagement and retention
improve even more.

There’s no limit to the ways you can incorporate creative

writing activities into your course… and you may find it’s
beneficial to establish objectives and parameters and then
require students to come up with their own ideas. A biomedical
student may decide to reimagine Mary Shelley’s Frankenstein
using today’s knowledge of genetic engineering. A health

101
professions student may write a journal from the point of view
of a person suffering from chronic illness. Virtually any
discipline could ask students to create characters in a specific
setting and time, build in backstory and a conflict, and engage
the characters in a scene where it’s critical that one of the
characters understands the concept or theory the other is
attempting to explain.

AI tools can assist students in brainstorming or transforming

their story ideas into more compelling forms. Jasper is an AI
writing tool that helps write long-form content and has over 50
storytelling tools, including a plot generator, character builder,
and scene creator. Students can also input their written dialogue
into ChatGPT and ask it to evaluate it through a series of lenses
(i.e., accuracy, entertainment value) and provide ideas for more
scenes or conflicts. Additionally, there are many AI programs
that generate images, allowing students to add visual elements
to their creative writing.

Sample ChatGPT prompts:

1. “Create a contemporary scene where Edward Jenner,

the inventor of the first vaccine (1796), returns to Earth
and attempts to explain how vaccines work to Andrew
Tate, a social influencer and anti-vaxxer. Generate
dialogue that incorporates the individual style and tone
of both Jenner and Tate.”
2. “Tom Hanks and Deadpool are running for Mayor and
campaigning on how they’ll reduce violent crimes.
Using a fast rap style, write a campaign speech for both.”
3. “Write a film treatment for a comedy about a
nutritionist who is hired by a famous Hollywood star to
help him get fit for his next role.”

102
 52
Generate: Software Coding
Provide students with the opportunity to
transform concepts and ideas into products and
solutions by using AI tools to generate software
codes.

People write software codes to create websites, apps, games, and

other programs that solve problems, automate tasks, manage
data, increase productivity, and entertain. When students are
instructed to turn course concepts into applicable, real-world
solutions using software coding, they engage more fully with
the content and see firsthand how their knowledge, ideas, and
creativity can positively impact themselves and others.

Traditionally, software coding was left to the engineers who

spent years studying its various programming languages, tools,
algorithms, and methodologies. Today, AI programs offer
advanced coding assistance, allowing students without
programming experience to experiment and engage in the
process.

Students can begin by clearly defining their app idea, its main
features, and how it solves a problem or otherwise benefits
users. They then choose a beginner-friendly programming
language like JavaScript and Python and ask ChatGPT to lead
them through the process. The free version of ChatGPT can

103
answer questions, explain concepts, brainstorm solutions,
translate different coding languages, and generate snippets of
coding that students can use to test their ideas on a smaller
scale. (The upgraded Plus version, GPT-4, can offer much more
advanced assistance in coding, if students want to continue to
pursue their project.)

Sample ChatGPT Prompts:

1. “Using this dataset, generate a software code that allows

me to analyze the relationship between income and
health in the United States.” [Paste text below].
2. “Show me how to write a software code for an app that
allows me to combine my digital art into a 3D collage.”

104
 53
Study: Summarize Long Articles
Encourage students to use ChatGPT to absorb
complex readings more fully.

As educators, we often worry that students will use AI tools

unethically, but ChatGPT also provides students with tools that
help them excel while doing their own work. One way is by its
ability to absorb large amounts of text, distill it into easily
comprehended language, and summarize its key points.

None of the ChatGPT versions have the ability to access or

browse external web content and are only “trained” on
information prior to 2021, but all versions will allow you to cut
and paste text into its “message” and request a summary of key
points. Currently, ChatGPT has a word limit of around 3,000
words while the paid version, GPT-4, has approximately a
6,000-word limit. Bing Chat—another AI tool that’s free for
personal use for Microsoft users—is integrated with other
applications and can search the internet with its engine. This
allows students to provide Bing Chat with a URL and direct it to
“read” and summarize the materials there. For PDFs that aren’t
online, students can upload them to a Google Drive to generate
a usable URL.

When prompting an AI tool to summarize, students can ask it to

lower its readability level, translate concepts into layperson’s
terms, or change the intended audience. They can also ask for

105
summaries with different word counts and compare differences.
If you assign this task, you may ask students to evaluate the AI-
generated summaries based on their interpretation of the
article’s key points.

Sample ChatGPT Prompts:

1. “Summarize this article as if you oppose its solutions.

Create 10 tweets with hashtags that include content
while making an argument.”
2. “Summarize the article, ‘Scientists propose a method
that imparts elastic recovery to ferroelectric materials’
into 20 bullet points. Write 10 bullets as if the reader is
a middle schooler and 10 bullets as if I’m a leader in this
industry.”
3. “Summarize this text as if it were a fairytale.”
4. “Summarize the causes and effects of climate change in a
table with two columns and 10 rows.”
5. “Annotate this article using APA formatting.”

106
 54
Study: Create Study Plans
Demonstrate how ChatGPT can help students
create study plans that include key course content.

Numerous studies have shown that many students don’t study

nearly as much as they should to fully learn a course’s content.
While college instructors typically recommend 25 hours a week
of study time, most students put in less than half of that time.

One way to encourage students to study more is to demonstrate

how to use AI tools to create study plans. ChatGPT has the
ability to summarize text, create outlines of terms and concepts,
and present information in other formats to meet students’
individual learning needs.

For example, a student studying an e-chapter on foundations of

early literacy, may cut and paste the material into ChatGPT and
ask it to create an outline of key points and real-life examples
for each key point.

Another student who’s struggling to retain topics or concepts

can ask ChatGPT to generate mnemonics, acronyms, rhymes,
stories, jokes, images, and analogies using different themes, such
as animals, colors, foods, TV shows, music, books, celebrities,
etc. to make the information more memorable.

107
Sample ChatGPT Prompts:

1. “Using ‘Friends’ references, create a mnemonic to

explain Newton’s Second Law of Motion.”
2. “Write three jokes that will help me remember the
Henderson-Hasselbalch Equation.”
3. “Summarize all of the key components of this chapter in
the form of a rap song.”

108
 55
Study: Question Creation
Encourage students to use AI tools to generate
questions for self-quizzing on course context.

Many students have grown up using Quizlet and similar apps

that host a large collection of flashcards and other study tools
created by students and educators. Students search that site for
study materials that have been uploaded by others.

ChatGPT can generate all types of quiz questions on any topic,

even if there are no existing materials, and it allows students to
customize the difficulty and style of the quiz questions. Students
can ask ChatGPT to create questions in a variety of forms,
including true/false, multiple choice, fill-in-the-blank, short
answers, and even multiple-step, real- and abstract-word
problems.

When students respond to ChatGPT questions, the program can

also provide instant feedback. Students can then ask follow-up
questions on the content and receive suggestions for further
reading.

109
Sample ChatGPT Prompts:

1. “Generate a 10-question quiz on this content. Include all

types of question formats, limit my time on the quiz,
and withhold the answers until I respond. Then,
evaluate my responses and offer suggestions for further
reading on questions I missed.”
2. “Generate a 10-point probability and statistics question
involving binomial distribution and standard deviation.”
3. “Create a matching question that tests the stages of
cognitive development based on Jean Piaget.”

110
 56
Study: Multiple-Choice
Question Generation
Communicate the effectiveness of using ChatGPT
to generate multiple-choice questions for self-
quizzes.

One of the most commonly used assessment tools is the

multiple-choice question. While they can be time-consuming to
create—especially when we’re focused on generating plausible
“wrong” answers—and are often limited to testing material
found on the lower levels of Bloom’s Taxonomy, they’re
frequently used for their ease of grading and ability to test large
amounts of content.

ChatGPT can assist students in generating all sorts of multiple-

choice questions that allow them to consider content in a
variety of ways. For more effective study tools, students may
find asking ChatGPT to provide more choices—up to six or
eight options—will better prepare them for assessments by
reducing their ability to determine the correct answer by using
elimination techniques. Having more choices can also
encourage students to think more critically.

For students who want to create a pretest, ChatGPT and other

tools can create quizzes that withhold answers until the

111
completion of the quiz and then provide feedback for missed
questions. Students can also prompt ChatGPT to create a
multitude of tests, varying the level of difficulty each time.

Sample ChatGPT prompts:

1. “Create an interactive quiz about the Central Limit

Theorem for a second-year college student with 5
questions. Start the quiz.” After the quiz, students can
follow up with this prompt: “Explain the content on
questions I answered incorrectly.”
2. “Generate a series of multiple-choice questions that
address influences and contexts of postmodern
literature. Provide 8 possible responses to each question.
Withhold answers until I respond.”

112
 57
Study: Find Elusive Terms
Demonstrate how ChatGPT can help students find
those terms they once knew, but temporarily
forgot.

All of us have experienced that moment of frustration when a

word, name, or term is just on the edge of our memory, but we
can’t quite retrieve it. This phenomenon is termed “tip-of-the-
tongue (TOT)” (or lethologica) and it defines that state when
we’re able to recall some features of the term—like the first
letter, the number of syllables, or a similar sound, word, or
meaning—but the rest of it remains elusive.

One of the first studies that looked at the TOT phenomenon was
conducted in 1966 by Roger Brown and David McNeill who
found this phenomenon occurred about once in every 50
questions and often the elusive term was one that was foreign,
long, or used infrequently.

Compared to most search engines, ChatGPT is a much more

effective tool for helping with this kind of recall and provides a
much more thorough response. By demonstrating to students
how to input an explanation for the term or concept they’re
attempting to retrieve, you’ll show them how to lessen that
frustration and find answers quickly.

113
Sample ChatGPT Prompts:

1. “Evolution is not a smooth curve. When graphed, the

curve takes turns plateauing and accelerating. What’s
this called?”
2. “What is the term for relapsing into criminal behavior?”
3. “What’s the fancy term for throwing someone out a
window?”

114
 58
Study: Personal Teaching
Assistant
Show students techniques they can use to double-
check course content with ChatGPT.

One of ChatGPT’s exceptional abilities is its ability to provide

explanations and definitions in a variety of ways. It shines at
this task like few others and provides vastly more useful results
than search engines.

Students can also request how they’d like to receive the content.
Depending on how they learn best, students can ask for charts
and graphs, images, parallax and 3D effects, audio recordings or
videos, text summaries in bullet forms, games or simulations,
and a variety of other modes.

Another ChatGPT feature that students may benefit from is its

ability to tailor content to accommodate different styles and
reading levels. Often, when learning an advanced concept, it’s
helpful to receive a series of descriptions that begin with
simplified explanations and progress in complexity.

115
Sample ChatGPT prompts:

1. “What is the function of classes in object-oriented

programming? Please provide a text-based explanation
and a parallax of the concept.”
2. “Define ‘economic equilibrium’ using a simile or
metaphor. Write one for each of these reading levels:
first grade, middle school, and college level.”

116
 59
Study: Flashcard Generation
Communicate the effectiveness of creating
customized flashcards with ChatGPT.

Flashcards are a popular and effective way of studying because

they promote active recall and help students move a concept or
definition from short-term to long-term memory. In one study,
researchers found that almost 70 percent of college students in a
psychology course used flashcards to study for exams. Students
who used flashcards for all of their exams scored significantly
higher than those who didn’t (Golding, Wasarhaly, & Fletcher,
2012).

ChatGPT can also assist students in creating flashcards by taking

the course content and using it to generate distilled
explanations, definitions, examples, and illustrations of key
concepts. Students can also use AI tools to turn content into a
question-and-answer format and create interactive quizzes. For
students whose second language is English, ChatGPT can also
translate content into a student’s native language, often
resulting in improved retention.

Some AI tools, like Bing Chat, can also take images (graphs,
diagrams, illustrations, etc.) and translate the content into text,
allowing students to better understand concepts described
visually. And most AI tools can provide real-world applications
to content which also boosts learning and retention.

117
Sample ChatGPT prompts:

1. “Using this content, provide step-by-step instructions

for creating an interactive quiz that utilizes flashcards.”
2. “Create visual flashcards for college-level, introduction
to biology course that is reviewing the structure and
function of cells.”
3. “Apply the theory of plate tectonics to real-world
scenarios in a series of 5 flashcards.”

118
 60
Study: Time Management
Help students boost productivity with time
management tools.

Numerous studies have shown that many college students suffer

from poor time management skills, resulting in lower grades and
an increased risk of dropping out. Even those who manage to
complete tasks after procrastinating experience higher levels of
stress and reduced sleep… two other factors that contribute to
poor student performance.

ChatGPT and other AI tools can become a student’s lifestyle

coach by creating schedules and other tools that help students
balance their academic, personal, and work lives. One way to
start is by encouraging students to ask ChatGPT to create a quiz
that assesses their time management skills. This first step is
often a wake-up call to students who aren’t always aware of
how much they’re juggling or that they lack these skills.

Once students determine their individual needs, ChatGPT can

create schedules that prioritize and organize their tasks, block
out time for those activities, set reminders, and help students
manage social media and other distractions. ChatGPT can also
provide students with strategies like the Pomodoro technique
which helps overcome procrastination.

119
Sample ChatGPT prompts:

1. “Create a weekly schedule that helps me manage my

class, study, work, and social time. I have [number] of
classes a week, each lasting [number] of hours. I want to
study for [number] of hours per class. I work [number]
of hours a week and would like [number] of hours a
week for fun or free time.”
2. “Generate SMART goals for improving my writing skills
this semester. Then break down goals into action steps
and provide methods on how to track my performance
and results.”
3. “I’m struggling to meet the goals I set in my time
management plan. Can you assess my productivity and
brainstorm solutions?”

120
 61
Study: Career Paths
Assist students in setting and reaching career goals
with ChatGPT tools that streamline the process.

It’s no secret that many college students are fearful of the future
and experience a great deal of anxiety when selecting a major or
career path. These fears—often coupled with accumulating
student debt—can create additional stress and frustration.

ChatGPT can offer self-assessment exercises that identify

strengths and interests and match those findings with a
multitude of career paths. It can also research industries and
their future outlooks, provide information on the job market
globally and in specific locations, and share details on job
responsibilities, growth opportunities, and salaries.

When students have determined a career path, ChatGPT can

provide information on what students need to do to secure those
jobs, including details on degree programs, certifications, and
specialized training. It can also offer tips on finding jobs from
networking ideas to interview questions. Finally, ChatGPT can
offer suggestions and provide examples of successful resumes
and other application materials.

121
Sample ChatGPT prompts:

1. “Describe a typical day in the life of someone working

in a [name] career. Include a daily routine, interactions,
and typical tasks.”
2. “Research the work environments typically associated
with a [name] career. Are there opportunities to work in
nonprofit organizations, corporate, or private
organizations? If so, what are the cultural aspects of
each sector? What types of personalities fit best in
each?”
3. Using Bing Chat, which has browsing capabilities,
prompt it to: “Generate a list of professional
development opportunities for this [name] career and
include events that are near [location]. Are there
opportunities to connect with mentors there? If so,
include the application process.”

122
 AI Glossary
Adobe Firefly – an image-generating AI created by Adobe
Sensei; trained only on licensed images and public domain
content
Bard – a text-generating AI created by Google using LaMDA
technology; interacts with the web
Bing Chat – a text-generating AI created by Microsoft based on
the GPT-4 model; interacts with the web
BlenderBot – a text-generating AI created by Meta; the latest
version (BlenderBot 3.0) interacts with the web
Canva – a “freemium” online image-creating/editing tool that
added AI-image generation to its paid subscription package
in 2023
ChatGPT – a text-generating AI chatbot created by OpenAI;
does not interact with the web
Claude – a text-generating AI created by Anthropic
(ex-employees of OpenAI); interacts with the web
CoPilot – an AI embedded in Microsoft Office products; can
generate codes
DALL·E – an image-generating AI created by OpenAI;
incorporated within Bing Chat
ErnieBot – a text-generating AI (in Mandarin) from Chinese
search engine Baidu
Falcon AI – a text-generating AI created by UAE’s Technology
Innovation Institute; transparent, open source, trained on
RefinedWeb, a custom-made dataset
LaMDA (Language Model for Dialogue Applications) – a LLM
trained specifically on dialogue, such as Google’s Bard

123
LLM (Large Language Model) – a type of software / AI that
accesses large databases it’s been trained on to predict the
next logical word in a sentence, given the task/question it’s
been given. Advanced models have excellent “perplexity”
(plausibility in the word choice) and “burstiness” (variation
of the sentences).
Jasper – a for-pay text-generating AI aimed at businesses and
blog posts
Midjourney – an image-generating AI; interacts with Discord
NightCafe – a free image-generating AI; uses Stable Diffusion
algorithms to produce low-resolution images
OpenAI – the company that created ChatGPT and DALL-E
Stable Diffusion – an image-generating AI created by the
startup Stability AI
Sydney – the name of the AI that supports Microsoft’s Bing
search engine
xAI – the ChatGPT competitor launched by Elon Musk

124
 About the Authors
Kevin Yee earned his Ph.D. in German Literature from UC
Irvine, and enjoyed teaching for several years as a full-time
faculty member at the University of Iowa and Duke University
before changing his focus to educational development when
joining the University of Central Florida in 2004. He is now the
director of UCF’s Faculty Center for Teaching and Learning
(https://fctl.ucf.edu).

Kirby Whittington formerly worked as the STEM Instructional

Specialist in UCF’s Faculty Center for Teaching and Learning.
She earned her Ph.D. in Curriculum and Instruction with a
focus on science teacher education from Florida State
University. She has worked as an educator, teacher educator,
researcher, and curriculum developer.

Erin Doggette is an educator of 15+ years with an earned Ph.D.

in Higher Education Administration & Leadership, an Ed.S. in
Curriculum & Instruction, and an M.A. and B.A. in
English/English Language Arts and Secondary Education. Her
passion for teaching and educational development led her to
join the UCF Faculty Center for Teaching and Learning as an
Instructional Specialist.

Laurie Uttich, a poet and prose writer, is a Senior Lecturer at the

University of Central Florida where she’s taught composition
and creative writing for 15 years. She is now an Instructional
Specialist at UCF’s Faculty Center for Teaching and Learning
and recently used Bing Chat to come up with names for her
son’s dog (result: Basho).

125
126