Learning & AI Minicourse

This is a sample module from a larger Learning and AI minicourse. It demonstrates an instructional approach for using generative AI to support varied, repeated practice while keeping students actively engaged in evaluating their learning.

Pattern Recognition with Variation

This activity works best for situations where students need to gain an understanding of nuance in a concept, nuance that often builds up with experience over time. This process takes advantage of repeated interactions with similar but varied problems.

When I was teaching introductory computer science classes to students, I often wondered how many exposures a student would need to a specific concept–for example, using loops to solve different types of math problems–before they would gain fluency with the topic. If it was a core topic, I would have been willing to create as many examples as necessary. But, not all students needed the same level of exposure to gain fluency or the same distribution of exposure to different aspects of the topic.

As an instructor, the challenge was to try to present enough examples that most students would feel comfortable with most of the concepts without boring the quick adopters or significantly under-serving other students. Having additional resources, such as text books or websites available with additional questions sometimes solved this, but not always.

Generative AI, with guidance, can add a deep level of support in this situation.

Select one of the topics below, preferably one that you know little or nothing about, and read through the explanations and the provided examples. Then paste the provided information into your preferred generative AI tool or use one of the tools the university provides.

Work the set of practice problems that it produces. Check your answers, then ask follow up questions, have it produce more questions about specific problems, or just wait and rework the problems after a spaced interval–whatever you need to feel comfortable with the topic.

You might also try explaining the patterns that you begin to recognize in the problems to the LLM to help articulate what you are discovering. Tomorrow, take the associated quiz for the topic in D2L. As always you will receive full credit for completing the quiz regardless of your score.

You are welcome to experiment with content from one of your own courses for this instead of trying one of the topics below. In that case, modify one of the prompts below to match a topic from your course. This may take some work as these prompts are more nuanced than those in Module 2. For the quiz, you will be asked to describe what you chose and answer a few questions about the process.

A: Northrop Frye's Theory of Archetypes

Overview

The following contains three parts. The first part is the overview of the topic, which is the same as the overview from Module 2. Then there is the prompt for this specific activity, followed by the sample questions.

You should review the notes, if necessary, and then try the practice problems. Then copy everything on this page except for this Overview section into your preferred generative AI tool or use one of the tools the university provides at https://genai.arizona.edu/, and start practicing.

In this activity, your goal is not simply to memorize the four mythoi. Through repeated encounters with different narratives, try to become more comfortable recognizing the characteristics that distinguish comedy, romance, tragedy, and irony/satire. As you work, pay particular attention to the overall movement of a narrative rather than individual plot events or whether a story seems happy, sad, funny, or serious.

Once you have the basic pattern, you could prompt the generative AI with something like, “In the next set of questions, use Hamlet, Star Wars, The Great Gatsby, Barbie, and other works I'm likely to know when possible.” Then, as you become more comfortable, try articulating the patterns you are recognizing. What makes you start thinking "tragedy" before reaching the end of a scenario? What distinguishes a triumphant romantic hero from a comic protagonist whose story ends in integration? What signals push an example toward irony/satire?


Northrop Frye’s Theory of Literary Archetypes

Notes

Brief Introduction

Literary critic Northrop Frye argued that individual stories are connected to larger patterns that recur across literature. In Anatomy of Criticism (1957), he described four broad narrative patterns, or mythoi: comedy, romance, tragedy, and irony/satire. Frye associated each of these patterns with a season, creating a cycle in which one form can blend into the forms that come before and after it.

The framework provides one way of thinking about what kind of story we are reading and how its overall movement resembles—or differs from—other stories.

Learning Outcomes

After reviewing this introduction and exploring the concept, you should be able to:

  • Identify the four major mythoi in Frye’s framework and their associated seasons.
  • Describe the general narrative movement and characteristics of each mythos.
  • Explain how neighboring mythoi overlap rather than functioning as completely separate categories.
  • Apply the framework to a familiar story and justify which mythos best describes it.

Description

Frye connects his four mythoi in a cycle:

Comedy → Romance → Tragedy → Irony/Satire → Comedy

He associates each with a season: comedy with spring, romance with summer, tragedy with autumn, and irony/satire with winter. The seasons represent the general movement of the stories rather than simply their settings.

Comedy (Spring) generally moves from disorder toward renewal and social integration. Characters encounter obstacles that prevent a desired outcome—often a marriage, relationship, or other form of reconciliation—but those obstacles are eventually overcome. The conclusion tends to create or restore a functioning community.

Romance (Summer) presents an idealized world of adventure, achievement, and heroic action. Its protagonists confront challenges or threatening forces and ultimately triumph over them. Romance represents the high point of Frye’s cycle: the hero and the world of the story can seem larger, more ordered, or more ideal than ordinary life.

Tragedy (Autumn) moves in the opposite direction. Rather than integration and triumph, the protagonist moves toward isolation, defeat, or death. The character may begin in a position of strength but becomes increasingly separated from society as the story progresses.

Irony and Satire (Winter) present a world in which characters have relatively little power. Instead of the heroic world of romance, we encounter limitation, absurdity, failed systems, or people trapped by circumstances they cannot overcome. Satire may expose the flaws of individuals or societies through ridicule.

These categories are not rigid boxes. Frye's cycle emphasizes their relationships. A comedy, for example, can move toward the idealized world of romance or toward the harsher world of irony and satire. Thinking about those boundaries can be more useful than simply assigning a story a label.

Prompt

I am practicing Northrop Frye's four literary mythoi through repeated encounters with varied examples. Use the provided overview, sample questions, and dimensions of variation as the basis for my practice.

Begin by generating 6 practice questions. Use the sample questions as models for their general structure, but create new narratives rather than minor variations of the examples. Include all four mythoi across the set and vary the difficulty. Start with relatively clear examples and gradually introduce more ambiguous cases, including narratives with characteristics of neighboring mythoi.

Do not show me the answers until I respond. After I answer, give me feedback on my reasoning as well as whether my answer is correct. If a question is genuinely ambiguous, acknowledge that and discuss what evidence supports different interpretations rather than pretending there is one obvious answer.

After the first set, let me decide what I need next. I may ask for more questions, practice with focus on a particular distinction, more difficult examples, examples using narratives I already know, or help articulating patterns I am beginning to recognize.

Sample Questions

These examples show the general structure and level of questions you should produce.

Example 1

A young man wants to marry a woman whose family disapproves of him. After a series of misunderstandings and conflicts, the families reconcile, the couple marries, and previously divided groups come together for a celebration.

Which of Frye's four mythoi best describes the overall narrative?

A. Comedy
B. Romance
C. Tragedy
D. Irony/Satire

Answer: A — Comedy. The important feature is not that the story is humorous, but its movement from conflict and obstruction toward reconciliation and social integration.

Example 2

A respected political leader begins the story at the height of his influence. His increasingly destructive decisions alienate his allies and family. By the end, he has lost his position, relationships, and eventually his life.

Which mythos best describes the narrative?

A. Comedy
B. Romance
C. Tragedy
D. Irony/Satire

Answer: C — Tragedy. The narrative moves from strength and social connection toward increasing isolation, loss, and defeat.

Example 3

A story ultimately ends with two estranged groups reconciling and forming a new community. However, much of the story uses absurd situations to expose the hypocrisy and incompetence of the institutions that created the conflict.

Which interpretation best applies Frye's framework?

A. It is comedy because stories can belong to only one mythos.
B. It is irony/satire because ridicule overrides the story's ending.
C. It primarily follows a comic movement but contains strong elements of irony/satire.
D. It cannot be analyzed using Frye because it contains characteristics of multiple mythoi.

Answer: C. Frye's categories are related rather than rigidly separated. A narrative can primarily follow one pattern while sharing characteristics with a neighboring mythos.

Important Dimensions of Variation

As you generate additional questions, vary more than the names and settings of the stories. Practice should include:

  • Clear examples of all four mythoi.
  • Different genres, settings, protagonists, and types of conflict.
  • Stories in which surface features could be misleading—for example, a funny tragedy or a comedy involving serious conflict.
  • Examples that require attention to the movement of the narrative: toward integration, triumph, isolation, powerlessness, etc.
  • Increasingly ambiguous examples that contain characteristics of neighboring mythoi.
  • Questions that ask you to explain or justify your classification rather than only select an answer.
  • When useful, examples drawn from books, films, television shows, or other narratives you already know.
B: Loops

Overview

The following contains three parts. The first part provides an introduction to basic for loops in Python. Then there is the prompt for this specific activity, followed by sample questions.

You should review the notes and then try the practice problems. Then copy everything on this page except for this Overview section into your preferred generative AI tool or use one of the tools the university provides at https://genai.arizona.edu/, and start practicing.

In this activity, your goal is to become comfortable reading simple loops and predicting what they will do. You do not need any previous programming experience. As you encounter different examples, pay attention to how small changes in the starting value, stopping value, direction, and statements inside the loop change the result.

As you become more comfortable, try articulating the patterns you are recognizing. Without tracing every step, can you begin to predict how many times a loop will run? What happens when the stopping value changes? When does a variable contain a running total rather than just the most recent value? Ask the generative AI tool to test the rules you think you have discovered with additional examples.


Understanding Basic for Loops

Notes

Brief Introduction

Computer programs frequently need to perform the same operation multiple times. A loop allows a programmer to repeat instructions without writing those instructions separately for every repetition.

In Python, a for loop can work with a sequence of numbers created using range(). By changing the values given to range() and the operations performed inside the loop, a programmer can control how many times an operation occurs and what happens during each repetition.

Learning Outcomes

After reviewing this introduction and exploring the concept, you should be able to:

  • Trace a basic for loop and predict its output.
  • Determine which values will be produced by different uses of range().
  • Explain how changes to the starting value, stopping value, and step affect a loop.
  • Trace loops that use a variable to accumulate a value across multiple iterations.
  • Recognize common boundary and off-by-one errors in simple loops.

Description

Consider this Python code:

for i in range(5):
    print(i)

The loop produces:

0
1
2
3
4

range(5) begins at 0 by default and stops before 5. The variable i takes each value in the sequence, and the indented statement runs once for each value.

We can specify a different starting point:

for i in range(2, 6):
    print(i)

This produces 2, 3, 4, 5. Again, the stopping value, 6, is not included.

A third value can specify the step:

for i in range(2, 10, 2):
    print(i)

This produces 2, 4, 6, 8. A negative step can make the values decrease:

for i in range(5, 0, -1):
    print(i)

which produces 5, 4, 3, 2, 1.

Loops can also update a variable during each iteration:

total = 0
for i in range(1, 5):
    total = total + i
print(total)

Here, total changes each time through the loop:

0 → 1 → 3 → 6 → 10

The final output is therefore 10.

Small changes to a loop can produce different results. Understanding those differences—especially where a range begins and ends, how many times a loop executes, and how values change during each iteration—is an important part of learning to read and write loops.

Prompt

I am practicing basic Python for loops through repeated encounters with varied examples. Use the provided notes, sample questions, and dimensions of variation as the basis for my practice.

Begin by generating 6 practice questions. Use the sample questions as models for their general structure, but create new loops rather than minor variations of the examples. Begin with relatively simple questions and gradually increase the difficulty. Vary the starting value, stopping value, step, number of iterations, and operations performed inside the loop. Include both questions that ask me to predict printed output and questions that ask me to determine the final value of a variable.

Do not show me the answers until I respond. After I answer, give me feedback on my reasoning as well as whether my answer is correct. If I make a mistake, help me identify the pattern I misunderstood rather than immediately giving me a lengthy explanation of loops.

After the first set, let me decide what I need next. I may ask for more questions, practice focused on a particular type of loop, more difficult examples, or help articulating patterns I am beginning to recognize.

Sample Questions

These examples show the general structure and level of questions you should produce.

Example 1

What is printed by the following code?

for i in range(2, 7):
    print(i)

A. 2 3 4 5 6
B. 2 3 4 5 6 7
C. 3 4 5 6
D. 3 4 5 6 7

Answer: A. The range begins at 2 and stops before 7.

Example 2

What is printed by the following code?

for i in range(10, 3, -2):
    print(i)

A. 10 8 6 4
B. 10 8 6 4 2
C. 10 9 8 7 6 5 4
D. 8 6 4

Answer: A. The loop begins at 10 and decreases by 2 each time. The stopping value of 3 is not included.

Example 3

What is the final value of total?

total = 0
for i in range(1, 5):
    total = total + i * 2

A. 10
B. 20
C. 30
D. 40

Answer: B — 20. The values of i are 1, 2, 3, and 4. The loop therefore adds 2, 4, 6, and 8 to total.

Important Dimensions of Variation

As you generate additional questions, vary more than the specific numbers used in the code. Practice should include:

  • Loops that begin at 0 and loops that begin at other values.
  • Different stopping values, emphasizing that the stopping value in range() is not included.
  • Positive and negative steps.
  • Steps larger than 1.
  • Questions asking how many times a loop executes.
  • Questions asking for the sequence of values assigned to the loop variable.
  • Questions asking what gets printed during each iteration.
  • Questions involving a variable that accumulates a sum or other value across iterations.
  • Examples in which small boundary changes produce different results.
  • Common off-by-one situations in which a loop executes one time too many or one time too few.
  • Questions that ask you to explain the pattern or predict the effect of changing one part of the loop, rather than only calculating an output.
C: Genetic Drift

Overview

The following contains three parts. The first part is the overview of the topic, which is the same as the overview from Module 2. Then there is the prompt for this specific activity, followed by the sample questions.

You should review the notes, if necessary, and then try the practice problems. Then copy everything on this page except for this Overview section into your preferred generative AI tool or use one of the tools the university provides, and start practicing.

In this activity, your goal is to develop an intuitive understanding of how chance, population size, and allele frequency interact through genetic drift. Rather than simply memorizing that genetic drift has a greater effect on small populations, use repeated examples to practice predicting what might happen under different conditions and distinguishing changes caused by genetic drift from those caused by natural selection.

As you become more comfortable, pay attention to the patterns you are beginning to recognize. What changes when the population becomes smaller? What can you predict about an allele and what remains unpredictable? What happens when a population recovers after a bottleneck? Try articulating some of these patterns to the generative AI tool and then testing your ideas against additional examples.


Genetic Drift: How Chance Changes Populations

Notes

Brief Introduction

Evolution does not occur only through natural selection. Sometimes the frequency of a genetic variant changes within a population simply because of chance. This process is called genetic drift.

Genetic drift occurs in populations of all sizes, but its effects are much stronger in small populations. Over generations, random changes can cause an allele to become more common, disappear entirely, or eventually become the only version of a gene remaining in a population.

Learning Outcomes

After reviewing this introduction and exploring the concept, you should be able to:

  • Explain how genetic drift can change allele frequencies within a population.
  • Distinguish genetic drift from natural selection.
  • Explain why genetic drift has a greater effect on small populations than on large populations.
  • Predict how chance events such as population bottlenecks can affect genetic variation.

Description

Imagine a population of 1,000 beetles in which half carry one version of a gene, or allele, and half carry another. Not every beetle will reproduce. By chance, the individuals that produce offspring might contain slightly more of one allele than the other. The next generation might therefore have a 51/49 distribution rather than 50/50.

Nothing necessarily made one allele more useful than the other. The change could simply be the result of which individuals happened to reproduce. This random change in allele frequency is genetic drift.

Now imagine the same situation with only ten beetles. If two beetles die before reproducing, their absence could substantially alter the genetic makeup of the next generation. Random events affecting two individuals out of 1,000 generally have a much smaller proportional effect than events affecting two out of ten. For this reason, genetic drift tends to produce larger fluctuations in small populations.

Over many generations, these random fluctuations can have important consequences. An allele can disappear completely from a population. Alternatively, one allele can reach fixation, meaning it becomes the only allele of that type remaining in the population. Both outcomes reduce genetic variation.

Genetic drift can become especially noticeable when a population suddenly becomes very small. A population bottleneck occurs when an event such as a natural disaster, disease, or habitat disruption dramatically reduces a population. The surviving individuals may represent only a random portion of the genetic variation that existed before the event. Even if the population later grows again, some of its earlier genetic diversity may already have been lost.

This distinguishes genetic drift from natural selection. With natural selection, differences in survival or reproduction are related to heritable traits that affect reproductive success in a particular environment. With genetic drift, allele frequencies change through chance. Both processes can operate in the same population at the same time.

The effects of genetic drift therefore depend not only on what genetic variation exists, but also on population size and chance. A variant does not have to be beneficial or harmful for its frequency to change across generations.

Prompt

I am practicing genetic drift through repeated encounters with varied examples. Use the provided overview, sample questions, and dimensions of variation as the basis for my practice.

Begin by generating 6 practice questions. Use the sample questions as models for their general structure, but create new scenarios rather than minor variations of the examples. Vary population sizes, allele frequencies, environmental events, and other relevant conditions. Include questions that require me to distinguish genetic drift from natural selection and questions that explore the effects of population bottlenecks. Start with relatively clear examples and gradually increase the difficulty.

Do not show me the answers until I respond. After I answer, give me feedback on my reasoning as well as whether my answer is correct. When an outcome cannot be predicted with certainty because genetic drift is random, make that uncertainty explicit rather than treating one possible outcome as inevitable.

After the first set, let me decide what I need next. I may ask for more questions, practice focused on a particular distinction, more difficult examples, or help articulating patterns I am beginning to recognize.

Sample Questions

These examples show the general structure and level of questions you should produce.

Example 1

Two isolated populations of the same species contain alleles A and B at approximately equal frequencies. Population X contains 40 individuals, while Population Y contains 4,000 individuals. Neither allele provides a reproductive advantage.

In which population would you expect genetic drift to have the greater effect on allele frequencies over the next several generations?

A. Population X
B. Population Y
C. The effect should be identical in both populations.
D. Genetic drift will not occur because neither allele provides an advantage.

Answer: A — Population X. Genetic drift occurs in populations of all sizes, but random differences in survival and reproduction have a proportionally greater effect in smaller populations.

Example 2

A population of insects contains green and brown individuals. Birds can see green insects more easily against the brown soil, causing green insects to be eaten more frequently before reproducing. Over many generations, the brown allele becomes increasingly common.

Which process best explains this change?

A. Genetic drift
B. Natural selection
C. Population bottleneck
D. Random fixation

Answer: B — Natural selection. The difference in reproductive success is associated with a heritable characteristic. The change therefore is not occurring solely through chance.

Example 3

A storm randomly kills 90% of an island population. Before the storm, several alleles were present in the population. The population eventually returns to its previous size.

Which statement best describes the likely result?

A. All of the original alleles will return as the population grows.
B. The population will necessarily become better adapted to the island.
C. Some genetic variation may have been permanently lost even though the population size recovered.
D. Genetic drift could not have occurred because the storm affected individuals randomly.

Answer: C. A severe population bottleneck can leave only a random subset of the population's previous genetic variation. Increasing the number of individuals afterward does not automatically restore alleles that were lost.

Important Dimensions of Variation

As you generate additional questions, vary more than the species or setting. Practice should include:

  • Populations ranging from very small to very large.
  • Different starting allele frequencies.
  • Comparisons between otherwise similar populations of different sizes.
  • Random changes in which individuals survive or reproduce.
  • Changes in survival or reproduction associated with advantageous or disadvantageous traits.
  • Population bottlenecks of different severity.
  • Populations that grow again after experiencing a bottleneck.
  • Situations in which an allele becomes more common, becomes less common, reaches fixation, or disappears.
  • Questions that distinguish what can be predicted probabilistically from what cannot be predicted because drift is random.
  • Questions that require explanation or prediction rather than only identification of a vocabulary term.

The goal of this activity is to provide students with the following when learning a new concept:

  • Encounters with multiple examples
  • Variations to develop deeper understanding
  • Opportunity for recognition of patterns in problems
  • Opportunity to articulate those patterns

Notice a few things about this approach. The instructor still needs to provide important information–what is the topic, what are important aspects to explore, what assessment questions will look like, and a couple examples.

The student takes responsibility for generating the questions, practicing with them, and assessing the quality of the questions. Generative AI can make mistakes, so students must be engaged with the process here. Ideally, as the instructor, you will do some pre-experimentation to check question quality for a topic and provide low stakes opportunities for students to assess their learning and correct misconceptions before a summative assessment.

Important Note: While I have found generative AI to be effective when generating short, targeted practice questions like the examples in this module, I have NOT found it to be as effective in generating more complex exam style questions. It often misses the nuance that a good instructor includes in and between those questions even when given sample exams written by the instructor.