Course Design

• Design and develop courses that make sense to students, ensure the course learning outcomes align to materials, activities and assessments (otherwise, students might find that the value of the work required for their learning is unclear to them).
• Make assignments and assessment tasks specific to your course and your students’ course-based experiences.
• Incorporate student reflection into assessments as part of each task’s required set of demonstrable critical thinking skills.
• Scaffold (and sequence) assignments to include project proposals/outlines, multiple drafts, annotated bibliographies.
• Adopt a portfolio approach: this will improve your ability to access not only examples of a student’s work but also gain a greater understanding of their learning process.
• Replace short answers, essays, and multiple choice tests with a variety of multimodal assessment tasks: in-person or virtual presentations, mind maps, podcasts, videos, recorded learning journals.
  • Anticipate and proactively design for the presence of students who will succeed more with accommodations
  • Prepare alternative assessments and assignments in advance
• Emphasize peer work such as peer editing and commentary as part of your student evaluation/writing process.

AIM: Use Generative AI to help develop a glossary for a course: Ask Generative AI to generate a glossary for your course, syllabus, unit, or lecture.

SUMMARY: Immersed in my field as I am, I often find myself using terminology that is unfamiliar to first-year students. Hence, I am concerned students will not follow my survey lecture on Japanese literature, so I want to give them a glossary in advance of my lecture, so they can be aware of the key terms I cover. I could input the following prompt into ChatGPT: “Can you generate a glossary of key terms for an [insert context] on [insert subject]?

COURSE CONTEXT: This type of prompt can be used for courses, lectures, unit overviews, and even conference papers.

EXAMPLE: An introductory lecture for a survey course on Japanese literature

• Prompt: Can you generate a glossary of key terms for an introductory course on Japanese literature?

• ChatGPT Output: Sure, here's a glossary of key terms for an introductory course on Japanese literature:

  1. Haiku: A traditional Japanese poem consisting of three lines. The first and third lines have five syllables each, while the second line has seven syllables.
  2. Tank̄a: A traditional Japanese poem consisting of five lines. The first and third lines have five syllables each, while the second, fourth, and fifth lines have seven syllables each.
  3. Kabuki: A traditional form of Japanese theater that originated in the Edo period. It is known for its elaborate makeup, costumes, and dramatic performances.
  4. Noh: A traditional form of Japanese theater that originated in the 14th century. It features sparse sets, simple costumes, and highly stylized movements.
  5. Kana: A Japanese writing system consisting of two syllabic scripts: hiragana and katakana. Kana is used alongside kanji (Chinese characters) to write Japanese.
  6. Kokoro: A Japanese word that can be translated as "heart," "mind," or "spirit." It is often used in Japanese literature to convey complex emotional states.
  7. Mono no aware: A Japanese concept that can be translated as "the pathos of things." It refers to a sensitivity to the impermanence of all things, and is often expressed in Japanese literature through depictions of fleeting beauty.
  8. Zuihitsu: A type of Japanese essay that is characterized by its informal style and lack of a clear structure. Zuihitsu often includes personal anecdotes, reflections, and observations.
  9. Meiji Restoration: A period in Japanese history from 1868 to 1912 when the country underwent significant political, economic, and social changes. The Meiji Restoration had a major impact on Japanese literature, as it led to the introduction of Western literary forms and styles.
  10. Shinto: The indigenous religion of Japan, which emphasizes the worship of nature and ancestors. Shinto beliefs and practices are often reflected in Japanese literature.

INSTRUCTOR TAKE-AWAYS: As you can see, ChatGPT provided a list of terms that instructors can use to inform their lecture. Centrally, the glossary makes instructors aware of certain terms that students may not be familiar with, so when designing and delivering their lecture, instructors can be sure to place special emphasis on certain terms. Instructors have a choice of providing a glossary to students to help students navigate the lectures or course content.

By employing their own subject area expertises, research, and experiences, instructors need to critically assess this glossary, augment the results, cut out what is not necessary, and/or add what is necessary according to the particular focal points of their lecture(s). For instance, if the definitions are too brief, instructors may need to add details or resources for further research, reading, and engagement.

Instructors can also decide to use this glossary as it is in an interactive way with students. For instance, instructors can provide students with this list as a “starting point” handout as a hard copy and/or via eClass, so as students move through the lesson and the course, the students can add details to these basic definitions.

Develop tasks that account for students of all abilities and understandings of course concepts (inspired by: CESE NSW “What Works Best in Practice).

• Example Prompt: Develop an assignment that offers a challenge for Sociology students who possess a strong grasp of ANOVA with Stata, while simultaneously providing adequate guidance and support for those who may be experiencing difficulties with the topic. Create a dataset that students could use.

AIM: Develop tasks that account for students of all abilities and understandings of course concepts (inspired by: CESE NSW “What Works Best in Practice).

SUMMARY: It can be time consuming to create well-thought-out assignments that take into account students of all abilities including disabilities. We can use Generative AI (ChatGPT) to do much of the heavy lifting for us by asking it to create an assignment for advanced skills and for those who are still working on a particular skill or topic. In the following example, I also ask ChatGPT to produce a dataset that students can use in their work. This can be particularly helpful to new instructors who have not gathered a toolkit of shareable data sources to provide students. This is also a quick way to provide new scenarios for students who want extra practice using ANOVA. In this example, I include a follow-up prompt to get the necessary output from ChatGPT.

COURSE CONTEXT: This is for an introductory Statistics course for the social sciences. The topic being covered is Analysis of Variance (ANOVA), a common analysis technique that is used in social sciences research.

• Example Starting Prompt: Develop an assignment that offers a challenge for Sociology students who possess a strong grasp of ANOVA using Stata, while simultaneously providing adequate guidance and support for those who may be experiencing difficulties with the topic.

• ChatGPT Output: Assignment Title: "Exploring Social Inequality through ANOVA Analysis in Stata"
  
  Background: Social inequality is a fundamental issue in sociology, and ANOVA is a useful tool to analyze and understand the relationships between various social factors and inequality. This assignment aims to challenge students to apply their ANOVA skills in Stata to investigate social inequality in a specific context.
  
  Instructions:

  1. Choose a dataset that includes variables related to social inequality (e.g., income, education, occupation, race, gender, etc.). You may use a publicly available dataset or one provided by your instructor.
  2. Create a research question that focuses on social inequality and select the appropriate variables to test your hypothesis.
  3. Conduct an ANOVA analysis using Stata to test your hypothesis.
  4. Interpret and report your findings, including any significant results, limitations, and implications for social inequality.
  5. Write a brief reflection on your experience conducting ANOVA analysis in Stata, including any challenges you faced and how you overcame them.

   
  Guidance and Support:

  1. Consult with your instructor and classmates to clarify any questions you may have about the assignment or ANOVA analysis in Stata.
  2. Review the course materials, readings, and tutorials related to ANOVA analysis in Stata.
  3. Seek assistance from the academic support services at your institution if you require additional support.
  4. Work collaboratively with your classmates, but ensure that your individual work is original and meets the assignment requirements.

   
  Assessment: Your assignment will be assessed based on the following criteria:

  • The clarity and coherence of your research question
  • The selection and appropriateness of your variables
  • The accuracy and completeness of your ANOVA analysis in Stata
  • The interpretation and reporting of your findings, including any significant results, limitations, and implications for social inequality
  • The quality and organization of your writing, including proper citation of sources and adherence to the assignment guidelines.

• FOLLOW-UP PROMPT: Create a dataset that students could use for their assignment.

  Sure, here is an example dataset that students could use for the "Exploring Social Inequality through ANOVA Analysis in Stata" assignment:

  Dataset Name: "Social Inequality in Education"

  Description: This dataset contains information on a sample of 500 students from a high school in a large urban area. The variables include gender, race, family income, parental education level, and student test scores.
  
  Variables:

  1. Gender: Categorical variable indicating the gender of the student (1 = Male, 2 = Female).
  2. Race: Categorical variable indicating the race of the student (1 = White, 2 = Black, 3 = Hispanic, 4 = Asian, 5 = Other).
  3. Family Income: Continuous variable indicating the total annual income of the student's family.
  4. Parental Education: Categorical variable indicating the education level of the student's parents (1 = Less than High School, 2 = High School, 3 = Some College, 4 = College Degree, 5 = Advanced Degree).
  5. Test Score: Continuous variable indicating the student's standardized test score.
     
     

     Sample Data:

     Gender	Race	Family Income	Parental Education	Test Score
     Male	White	40,000	High School	85
     Female	Black	25,000	Less than HS	65
     ...	...	...	...	...

     Note: This dataset is for example purposes only, and the data is entirely fictional.

INSTRUCTOR TAKE-AWAYS: While the output from ChatGPT may need to be modified to best fit your needs, the output provides a fast way to generate a framework to build assignments in your courses.

• Quick Prompt 1 (for Instructors): Generate lesson plans that are based on the skills and needs of your students (inspired by: CESE NSW “What Works Best in Practice).

  • ChatGPT: Example Prompt: Create a lesson plan for a 100-level university Engineering Mechanics course on kinetics that discusses the position, velocity, and acceleration of particles and rigid bodies. Include a variety of activities and alternative assessments.

• Quick Prompt 2 (for Instructors): (Suggestion inspired by an Ethan Mollick Tweet) Generate Harvard Business School styled case study on ‘original’ fictional case crafted by the instructor. The instructor can direct ChatGPT to incorporate real references and information. This is also a great way for instructors to use AI to quickly generate contemporary versions of classic HBS cases.

  • ChatGPT/GPT-4: Example Prompt: You are a business school professor and need to write a Harvard Business School case. You’ve chosen the topic of Google managing AI, when subject to Arrow’s cannibalization problem and the innovator’s dilemma.
  • Include additional steps (or Mollick’s chain of thought) in the starting prompt to ensure the chatbot’s output contains the details you need. Example below:

Additional Steps (Chain of thought):

Step 1. Consider how these concepts relate to Google.

Step 2: Write a case that revolves around a dilemma at Google about releasing a Generative AI system that could compete with search.

Step 3: Write an instructor’s note. Please provide a detailed summary of your insight and reevaluate it for improvement. You will then take the opposite position and produce a compelling rebuttal.

Title: Google's Dilemma: Managing Generative AI in the Face of Cannibalization Problem and Innovator's Dilemma

* See below suggestion, Link essay to novel case study, for more ideas.

• Quick Prompt 3 (for Instructors): Generate assignments that ask students to use ChatGPT. For such assignments, instructors should include a reflective and/or evaluative component which asks students to reflect on their use of ChatGPT and/or critically evaluate the responses from or exchanges with ChatGPT.

  • Sample Prompt: Create an assignment that asks [Indicate Course] students to use ChatGPT.
  • Example Prompt: Create an assignment that asks History students to use ChatGPT.

• Quick Prompt 4 (for Instructors): Generate activities that ask students to use ChatGPT. For such assignments, instructors should consider including a reflective and/or evaluative component which asks students to reflect on their use of ChatGPT and/or critically evaluate the responses from or exchanges with ChatGPT.

  • Sample Prompt: Create an assignment that asks [Indicate Course] students to use ChatGPT.
  • Example Prompt: Create an assignment that asks History students to use ChatGPT.

When students use AI tools in your courses, try to add reflection components to their course work (in-person or hybrid in-class activity or assignment/assessment). Create opportunities (have ChatGPT generate model responses or statements) for students to interrogate the output(s). This will help them learn its capabilities and limitations in a variety of contexts and applications. For instance, design problems that encourage students to critically analyze the relevance and factual accuracy of its responses (ask if fabrications are present). Also ask students to reflect on their own learning. How did they use the AI tool and how did it impact their learning?

For example, consider developing a straightforward Think-Pair-Share Problem Solving activity: divide your class into two groups (ChatGPT-users and Non-ChatGPT-users). Give them a problem (with multiple solutions) to engage with and solve. Consider an issue of particular relevance to your discipline. After an appropriate amount of time has passed, bring the two groups back together to discuss their various solutions. Each pair can present their response, explaining how they arrived at the solution and how they have and haven’t relied on ChatGPT. Ask them what they learned about working with their peers and with/without ChatGPT. (Inspired by Chowdury and Samuel, see Suggested Resources)

Avoid ready-made materials already found on the web. Link your writing assignments (essays) to unique case studies (see how to do this using ChatGPT in the above Quick Prompt example). Or, you can create case studies yourself by mixing and merging details from various existing source materials to develop novel situations and scenarios to share with your students. Using case studies in this way can encourage critical analysis, application, and evaluation of concepts, while fostering independent argument construction. (Inspired by Chowdury and Samuel, see Suggested Resources)

Before AI, you may have built assignments and rubrics yourself. Now you can do that with the help of AI. For instance, instead of building a rubric from scratch, collaborate with AI.