With the onset of AI and its increased availability, it has become a hot topic in higher education. In their 2023 literature review, Pradana et al. synthesized 93 articles about the use of ChatGPT in education, highlighting its increasingly profound popularity, since ChatGPT did not become available until 2021. While most of this conversation is centered around students’ use of AI in the classroom, less has been said about how faculty might benefit from AI in their teaching. One thing the research does seem to agree on is that AI and Chat GPT are good at reorganizing existing information and are less good at making up new information. For instance, Kiryakova and Angelova (2023) explored Bulgarian university professors’ perspectives on the potential and challenges of integrating ChatGPT into teaching activities. Their survey results indicated overall positive attitudes toward using ChatGPT to support teaching tasks, engage learners, and promote critical thinking (Kiryakova & Angelova, 2023). Therefore, this chapter focuses on how instructors use AI in higher education teaching. Instructors primarily use AI to help them develop content in two ways: to create content and to deliver content.

On using AI to develop content

Instructors can use AI tools to generate or curate educational content, such as quizzes, assignments, rubrics, lecture materials, presentation slides, and reading lists. They may also use it to create brief summaries of assigned course materials, module or course overviews and recaps, and even to reply to student messages and give feedback (Crompton & Burke, 2023; Joyce, 2023; LaPierre, 2023; Mondal et al., 2023). Meron and Araci (2023) looked at ChatGPT’s potential to help instructors design higher education courses. In their article, they used ChatGPT to develop materials and then reflected on what that process looked like by evaluating the strengths and weaknesses they encountered. They noted that ChatGPT was great for saving time, structuring text, and aiding brainstorming, its weaknesses included generating generic content requiring substantial human intervention and refinement (Meron & Araci, 2023).

On using AI to deliver content

AI can also be used to help instructors deliver content to students, and this is particularly the case in online courses. Instructors might use text-to-speech to create a voiceover with an AI-generated voice, they might use an avatar instead of their real image, or they might use an AI rendering to create video and/or audio.

In terms of using AI to create audio content, research so far has focused on the quality of the audio and how humanlike it sounds. For instance, Kim et al. (2022) found that AI instructors with humanlike voices are seen as more credible than those with machinelike voices. Additionally, the perceived social presence mediates this relationship, and credibility influences students’ intentions to enroll in future AI-led courses. This highlights the significance of creating AI instructors that students perceive as credible.

In terms of using AI to create video content, research so far has focused on whether instructors include themselves in the video versus an avatar of some kind. For instance, Yuan et al. (2021) explored how changes in instructors’ image and voice in online video courses impact learner outcomes. Two experiments were conducted with college students, altering instructor images and voices. Results showed that videos without instructor images led to better academic performance and less boredom. Real instructor images enhanced satisfaction with interaction and teaching. Real images were preferred over virtual ones, and beautified images over original ones. Real voices were preferred over synthesized ones. Positive evaluations of instructor image and voice correlated with social presence, satisfaction, enjoyment, and negatively with boredom. Instructor image positively predicted transfer learning achievement. The study suggests integrating instructor presence with course design and utilizing technology for image and voice processing in online course development.

Finally, in terms of using AI to create audio and video content,

For instance, Pi et al. (2022)  explored the effectiveness of using AI-generated virtual instructors in video lectures as a cost-effective alternative to real human instructors. Their results indicated that the AI likable instructor significantly improved students’ learning performance and motivation without diminishing their attention to the learning materials. This suggests that instructional designers can leverage AI voices and images of likable humans to enhance student motivation and learning outcomes in instructional videos.

Best Practices for Ethically Using AI to Create or Develop Content

Almost every study also considers, to varying extents, the ethics of using AI to develop and deliver content. However, almost all of these considerations are rather vague statements indicating that more research is necessary. In this section, we summarize these vague statements into a couple of best practices that we use and hope that others can use as they develop content with AI. Based on this, there seem to be two main best practices for integrating AI:

1. always review AI-generated content for accuracy and quality
2. always cite any AI tools used to help make content

The first best practice is to cite the AI tools or resources employed, similar to how you would cite any other source or technology used in your research or educational materials. There are several ways you could do this, depending on the AI and how it is used. You could:

In-Text Citations: One option would be to include the name of the AI tool or technology and the year of its publication or development in parentheses within the text where it is first mentioned. For example: “This content was drafted  using OpenAI’s ChatGPT-3 model (OpenAI, 2024).” You could put these in-text citations in everything in the course that you use AI for, including paragraphs of text, videos, slides, etc. This might be a good option if you regularly used AI throughout the course to develop content.

Reference List: Another option would be to include a full citation in the reference list at the end of your course materials. The citation should include the author (if applicable), the name of the AI tool or technology, the year it was developed or published, and any other relevant information. For example: “OpenAI. (2024). ChatGPT-3.” You could even link to the exact chat you used for full transparency.

Course Design Acknowledgements: Finally, you may also include a more general statement in your course that specifies the technologies you used in the course and how you used them. For example, you might say something such as “Some of the content in this course was drafted or refined using AI (Open AI, ChatGPT version 3.5). All content has been reviewed and meets the university, instructor, and instructional design team’s standards for quality.” This might be a good option if you primarily used AI to refine your existing content.

Related to these best practices, you may also consider including a notice about your expectations and preferences for students using in AI in your course by clarifying whether students are permitted to use AI for the assignments in the course. For example, you might include a statement like

In this course, you are encouraged to use AI tools to support your work. To maintain academic integrity, if you use AI tools, you must disclose that you used the AI tool and properly attribute it using in-text citations and quotations and listing it on the reference page. This should include the name(s) of the tool(s) used and a note about how the tool was used for the assignment.

Whatever you choose, by citing AI tools and resources properly, you give credit to the developers and provide transparency to your audience about the sources and technologies used in your course materials. Further, citing the AI materials used in your course, you demonstrate to students the importance of academic integrity and proper attribution in academic work. This modeling encourages students to follow ethical practices in their own research and writing, fostering a culture of honesty and respect for intellectual property. It also helps students understand the significance of giving credit to the creators of the AI tools and resources they rely on, promoting responsible scholarship and integrity in their academic endeavors.

Second, always verify the accuracy and quality of AI-generated content before incorporating it into the course. Review the material for factual errors, clarity, and coherence to maintain credibility and promote effective learning. There are also several best practices for doing this, including:

Read, listen, and watch any content that you used AI to make. Verify the accuracy of information presented in the content. AI may sometimes generate inaccurate or outdated information, so it is essential to fact-check and ensure that the content aligns with current knowledge and standards in the field. You never know when it might decide to throw something completely random, off-base, or inappropriate out there.

Verify that the content is original and not plagiarized from other sources. AI-generated content should be unique and add value to the course rather than simply replicating existing materials.

Verify the answers to multiple-choice questions generated with AI. Even if you gave it the exact text to create the questions, be sure that there is only one correct answer and that it is actually the correct answer.

By following these best practices, educators can harness the power of AI to create high-quality, engaging, and effective content for online higher education courses while supporting student learning and success.

Teaching Contexts for AI Use

Quizzes

Using ChatGPT to create effective online quizzes and quiz questions involves a few key steps. Here’s a guide to help you get started:

1. Identify what is being Assessed

• Determine the key concepts, skills, or knowledge areas your quiz will cover.
• Align questions with these objectives to ensure the quiz effectively measures the desired outcomes.

2. Decide on Question Types

• Multiple Choice: Good for assessing recognition and recall.
• True/False: Useful for checking basic understanding.
• Short Answer: Tests recall and comprehension.
• Fill-in-the-Blanks: Measures understanding of key terms or concepts.
• Matching: Helps assess the ability to connect related concepts.

3. Prompt ChatGPT to Create Questions

• Provide context and specify the type of questions you need.
• Example prompts:
  • “Generate a multiple-choice question about the causes of World War II.”
  • “Create a true/false question on the fundamentals of photosynthesis.”
  • “Provide a short answer question about the water cycle.”

4. Ensure Quality and Accuracy

• Review the questions for clarity and correctness.
• Make sure there’s only one correct answer for multiple-choice questions and that the options are not misleading.

5. Create Explanations

• Generate feedback or explanations for each question to enhance learning.
• Example prompt: “Provide an explanation for why the answer to the following multiple-choice question is correct.”

6. Test

• Test your quiz with a small group to identify any issues.
• Use feedback to refine questions and improve clarity.

Example Interaction with ChatGPT

Prompt: “I need a multiple-choice question on the basics of machine learning, including four answer options with one correct answer.”

ChatGPT Response:
Question: What is the primary purpose of supervised learning in machine learning?

1. To find hidden patterns in data
2. To make predictions based on labeled data
3. To group similar items without prior labels
4. To evaluate the performance of a model

Answer: b) To make predictions based on labeled data

Feedback: Supervised learning involves training a model on a labeled dataset, which means the data includes input-output pairs. The goal is to make predictions or decisions based on this labeled data.

Using these steps and examples should help you effectively create and refine online quizzes with the assistance of ChatGPT.

OpenAI. (2024). How to use ChatGPT to make quizzes. ChatGPT 3.5.

Assignments

Discussions

If you’re thinking that all of that is too much to do, you’re not on your own. You can use AI to help you make your discussions good. For instance, Chat GPT is a great tool to help you achieve these discussion best practices. For example, you may ask Chat GPT to . . .

• Can you suggest a follow-up prompt for this discussion so students know how to reply?
• Can you suggest follow-up questions to extend the discussion based on this prompt?
• How can I add depth and complexity to this discussion prompt?
• What modifications can I make to encourage critical thinking in responses to this prompt?
• What can I add to this prompt to encourage reflective responses?
• What can I do to prompt consideration of alternative viewpoints in response to this question?
• What multimedia elements can I add to enhance this discussion prompt?
• How can I connect this discussion prompt to real-world applications?

Examples

Okay so let’s take a look at some examples.

Example 1

Read this prompt and think about how you or your students might reply to it.

Original

Okay, now think about how we could revise it to make it better.

Revised

The revised prompt introduces a practical scenario (garden gnome theft) for relatability, enhancing engagement. It incorporates a second part that encourages interaction, critical thinking, and the application of class concepts when analyzing a classmate’s post, thereby deepening the discussion and promoting a comprehensive understanding.

Example 2

Okay, now let’s take a look at another example.

Original

Revised

The revised prompt gives background on the reading, improves structure and expectations for replies, and prompts students to share personal experiences, adding a practical and relatable aspect to the discussion.

Rubrics

Making good rubrics with specific and detailed criteria for each rating can be difficult and time-consuming. Enter—Chat GPT. Chat GPT can transform existing assignments and discussion prompts into detailed and specific rubrics. Chat GPT is great at restructuring information (Estrada & Fabián, 2024), so it’s a great way to make a rubric if you already have the assignment directions made. Such an approach expedites the rubric-making process and ensures that the rubric aligns well with the existing assessment. Chat GPT can be very helpful in developing assessment rubrics (Estrada & Fabián, 2024). Estrada and Fabián (2024) found that the AI-generated rubric in their study provided a clear and transparent framework for evaluating student work. So, Chat GPT allows instructors to efficiently generate rubrics, so long as the goal is to use the generated rubric, or parts of it, that align with the goals and outcomes of the assessment (Estrada & Fabián, 2024). While instruments are not fully replicable due to Chat GPT’s unique responses in each interaction, their responses can serve as a great starting point for developing rubrics. Estrada and Fabián (2024) recommend developing a clear and specific prompt in the message sent to Chat GPT. The directions should include instructions for what you want Chat GPT to do, the format you want the output to be in, and any other restrictions you have for the rubric (Estrada & Fabián, 2024). As with anything generated with AI, they noted the importance of reviewing the generated content for accuracy.

How To Make A Rubric With Chat GPT

Step 1: Locate the assessment you want to make a rubric for.

Before you can use Chat GPT to make a rubric, you need to have your assignment or rubric.

For the purposes of this example, we’re using the example in Box 1. The assignment directions outlined in Box 1 are from a graduate course in assessment and evaluation. The assignment is for students to write a 1-2 page reading reaction where they engage with the materials assigned in that module.

Box 1: Reading Reaction Assignment

Directions

After reading the assigned materials for the module, you will write a two-page (or one-page single-spaced) paper to react to the assigned readings in this module. Do not simply summarize the reading. Instead, engage the content by agreeing, disagreeing, applying, questioning, and/or extending the course materials. In doing so, you may write about anything that stood out to you as valuable in the materials. I’ve included questions below to get you thinking about the material more, but you can reply to some or none of these.

• How do diverse assessment methods like simulations and e-portfolios enhance learning engagement and authenticity?
• What are the benefits and challenges of integrating peer assessment and multimedia in educational settings?
• Reflect on how technology can be effectively utilized to improve assessment practices in your own teaching or learning context.
• What are the key factors influencing the high-stakes nature of assessment in higher education?
• How can educators balance the pressures of quality assurance, employability needs, and diversity considerations in assessment practices?
• Discuss strategies or reforms that could potentially alleviate challenges such as grade inflation and student dissatisfaction.
• Compare and contrast formative and summative assessment approaches. How can each type contribute to enhancing student learning outcomes?
• Reflect on how assessment data can be effectively used to inform instructional decisions and program improvements.
• Discuss the importance of aligning assessment strategies with educational goals and student needs.
• Why is it important for assignments to resonate culturally and hold significance for students? How can instructors achieve this in diverse educational settings?
• Reflect on your own experiences with culturally affirming assignments. How did they impact your learning and engagement?
• Discuss practical strategies for communicating course design and assignment expectations effectively through the syllabus.
• How are you doing? Is there anything you’d like me to know or share?

Guidelines

• Refer to the rubric for additional grading criteria.
• Do not simply summarize the reading. Instead, engage the content by agreeing, disagreeing, applying, questioning, and/or extending the course materials.
• Cite any additional materials used in APA format.

Step 2: Decide if you have any specific criteria and ratings in mind.

Criteria

You don’t need to know exactly what your criteria will be, but it does help to have a general idea of what they should be. Ideally, the criteria for the rubric should align with the assessment and the learning outcomes associated with that assessment (Chen & Ho, 2019). Generally speaking, you should aim to have about three to nine criteria for any given assessment. Unless you tell it to give you fewer, Chat GPT will often give you a lot of criteria, so even if you don’t have something specific in mind you still might want to specify the number of criteria you’re looking for.

Ratings

As Chan and Ho (2019) describe, writing the descriptions for each rating is the most challenging part of creating rubrics (p. 542). This is where Chat GPT really shines when you’re using it to help you make rubrics. Again, if you want all your rubrics to use a consistent rating scale, it would be good to tell Chat GPT which one you want to use, or, again, at the very least, the number of ratings you want to have. Some common ratings include:

• Met and not met or marks and no marks
• Meets expectations, approaching expectations, does not meet expectations (this is my favorite)
• Excellent, Average, Poor
• A, B, C, D, F or 5, 4, 3, 2, 1

In any case, you will want to include an explanation of the expectations for each rating, including a description of the characteristics of that level of work. This is where Chat GPT comes in.

Step 3: Write and enter your Chat GPT prompt

Now that you have your prompt and any criteria or ratings identified, you can enter your prompt into Chat GPT using the following template:

“Can you make a rubric in table format with the following ratings [insert ratings you selected in step 3] and with these criteria [insert criteria you identified, if any, in step 2] for this assignment: [copy and paste assignment directions]?”

For this example, I used this prompt: Can you make a rubric in table format that uses the following ratings: meets expectations, approaching expectations, and does not meet expectations; and the following criteria: content, examples, and writing quality; for this assignment: [here I copy and pasted the entire assignment directions from Box 1]?

The complete response from Chat GPT is provided in Box 2.

Box 2: Chat GPT’s Rubric

Absolutely! Here’s a rubric for the Reading Reaction Assignment using the ratings “Meets Expectations,” “Approaching Expectations,” and “Does Not Meet Expectations”:

Reading Reaction Assignment Rubric

This streamlined rubric should provide clear guidance on the key aspects you want to assess in the Reading Reaction Assignment.

Note: The information in this box was generated completely by Chat GPT 3.5 using the previously disclosed prompt.

Step 4: Regenerate, revise, and make your final rubric

Then, I used the parts of Chat GPT’s rubric that I liked to finalize the rubric for the assignment, which is provided in the table below. Be sure to also assign point values that match the point values in your assignment and syllabus, since sometimes Chat GPT likes to put random values in there for you. Keep in mind that the point value you assign to each criteria helps highlight the priority or importance of each (Chen & Ho, 2019).

For the rubric in Box 3, most of it looks pretty similar to the rubric Chat GPT gave me in Box 2. I really only summarized or combined pieces of information to make it a little bit more workable to the assignment and my preferences.

Box 3: Final Edited Rubric

Note: The information in this box is based on the content generated by Chat GPT 3.5 in Box 2.

Once you have everything just the way you want it, you can copy and paste the rubric into Canvas or whatever LMS you’re using. Refer to the links below for help adding a rubric to your assignments and discussions in Canvas.

• How to add a rubric to an assignment in Canvas
• How to add a rubric to a discussion in Canvas

Then, as always, be sure to test out the rubric. If you can, you can use a previous student’s assignment to make sure the rubric works the way you want it to before actually using it in a course. After you use it in a course, consider how well students met the criteria and whether it was easy to use it to grade the assignment; you want to make sure that the rubric aligns well with the assignment, outcomes, and the quality of work you get from students. As you use the rubric for grading, be sure to note anything you didn’t like or would want to change for next time.

• If the rubric is resulting in unexpected grades, you may need to rebalance the number of points for the criteria to better align with your expectations.
• If you consistently scored students lower on specific criteria, consider whether you can make the wording clearer or more objective.
• If you had to frequently adjust how you interpreted the rubric, you may need to revise your criteria to better match the outcomes associated with the assignment.
• If a criteria was consistently difficult to grade, you might need to add more criteria or split one criteria into multiple so you can describe them with a bit more detail.

Remember that regularly reflecting on and revising your rubrics helps make sure it is doing what it’s supposed to be doing and is helpful for both instructors and students.

Feedback

Using ChatGPT to provide students with feedback can be a great way to offer personalized and constructive comments. Here’s how you can effectively utilize ChatGPT for this purpose:

1. Identify what feedback is needed

• Determine whether you need feedback on specific assignments, quizzes, essays, or general performance.
• Consider the level of detail and the tone appropriate for your students.

2. Generate feedback by providing context and specifics

• Input the student’s response or work into ChatGPT along with any relevant criteria or guidelines.
• Example prompts:
  • “Here is a student’s essay on climate change. Provide feedback focusing on clarity, structure, and argumentation.”
  • “This is a student’s answer to a math problem. Give feedback on accuracy and problem-solving approach.”

Example Interaction:

Student Submission: “The causes of climate change are natural and man-made. Natural causes include volcanic eruptions and variations in solar radiation. Human activities such as burning fossil fuels contribute to climate change.”

Prompt for ChatGPT: “Provide feedback on the following student answer about climate change, focusing on completeness and accuracy.”

ChatGPT Response: Feedback: “Your answer provides a good overview of climate change causes. You correctly mention both natural and human-induced factors. To strengthen your response, consider elaborating on how human activities contribute specifically to greenhouse gas emissions and their impact on global temperatures. Additionally, providing examples of recent studies or data could enhance the depth of your analysis.”

3. Personalize the Feedback

• Customize the feedback to address individual strengths and areas for improvement.
• Use student names and refer to specific aspects of their work to make feedback more personal.

4. Encourage improvement by offering actionable suggestions

• Include clear, actionable steps for improvement.
• Example prompt: “Provide suggestions for improving this student’s essay on the impact of technology on education.”

ChatGPT Response: Suggestions: “To improve your essay, consider adding more examples of technological advancements and their specific impacts on educational practices. Additionally, addressing counterarguments and discussing potential drawbacks of technology in education would provide a more balanced perspective.”

5. Provide encouragement by maintaining a positive and supportive tone

• Acknowledge the effort and progress the student has made.
• Example prompt: “Generate a positive and encouraging comment for a student who has shown improvement in their writing skills.”

ChatGPT Response: Encouragement: “Great job on your recent writing assignment! I noticed significant improvement in your use of descriptive language and organization. Keep up the good work, and continue to focus on developing your ideas further. You’re making excellent progress!”

6. Integrate Feedback into Learning

• Suggest resources or activities that can help the student improve in specific areas.
• Example prompt: “Suggest resources or exercises for a student who needs help with essay structure.”

ChatGPT Response: Resources: “For improving essay structure, you might find these resources helpful: ‘How to Write an Effective Essay’ by [Author], available at [Link], and practice exercises on essay organization at [Website]. Additionally, reviewing sample essays and analyzing their structure can provide valuable insights.”

By following these steps, you can use ChatGPT to provide detailed, personalized, and constructive feedback that helps students understand their strengths and areas for improvement.

OpenAI. (2024). How to use ChatGPT to provide feedback. ChatGPT 3.5.

References

Andrade, H. G. (2005). Teaching With Rubrics: The Good, the Bad, and the Ugly. College Teaching, 53(1), 27–31. https://doi.org/10.3200/CTCH.53.1.27-31

Brookhart, S. M., & Chen, F. (2015). The quality and effectiveness of descriptive rubrics. Educational Review, 67(3), 343–368. https://doi.org/10.1080/00131911.2014.929565

Chan, Z., & Ho, S. (2019). Good and bad practices in rubrics: The perspectives of students and educators. Assessment & Evaluation in Higher Education, 44(4), 533–545. https://doi.org/10.1080/02602938.2018.1522528

Crompton, H., & Burke, D. (2023). Artificial intelligence in higher education: The state of the field. International Journal of Educational Technology in Higher Education, 20(1), 22. https://doi.org/10.1186/s41239-023-00392-8

Estrada, B. O., & Fabián, A. D. H. (2024). Generación de rúbricas con herramientas de inteligencia artificial para la evaluación de aprendizajes en educación superior. DIDAC, 84, Article 84. https://doi.org/10.48102/didac.2024..84_JUL-DIC.211

Gezie, A., Khaja, K., Chang, V. N., Adamek, M. E., & Johnsen, M. B. (2012). Rubrics as a Tool for Learning and Assessment: What Do Baccalaureate Students Think? Journal of Teaching in Social Work, 32(4), 421–437. https://doi.org/10.1080/08841233.2012.705240

Jonsson, A. (2014). Rubrics as a way of providing transparency in assessment. Assessment & Evaluation in Higher Education, 39(7), 840–852. https://doi.org/10.1080/02602938.2013.875117

Joyce, A. (2023). Embracing AI: Using Cat-GPT to Encourage Classroom Discussion. College Teaching, 0(0), 1–3.

Kiryakova, G., & Angelova, N. (2023). ChatGPT: A Challenging Tool for the University Professors in Their Teaching Practice. Education Sciences, 13(10), NA-NA. https://doi.org/10.3390/educsci13101056

LaPierre, S. S. (2023). Educators Find Value in ChatGPT. Computers in Libraries, 43(7), 31–35.

Lowenthal, P. R. (2009). The evolution and influence of social presence theory in online learning. Social Computing, Concepts, Methodologies, Tools and Applications, 113–128.

Meron, Y., & Araci, Y. T. (2023). Artificial intelligence in design education: Evaluating ChatGPT as a virtual colleague for post-graduate course development. Design Science, 9, e30. https://doi.org/10.1017/dsj.2023.28

Mondal, H., Marndi, G., Behera, J., & Mondal, S. (2023). ChatGPT for teachers: Practical examples for utilizing artificial intelligence for educational purposes. Indian Journal of Vascular and Endovascular Surgery, 10(3), 200–200.

Pi, Z., Deng, L., Wang, X., Guo, P., Xu, T., & Zhou, Y. (2022). The influences of a virtual instructor’s voice and appearance on learning from video lectures. Journal of Computer Assisted Learning, 38(6), 1703–1713. https://doi.org/10.1111/jcal.12704

Picciano, A. (2002). Beyond student perceptions: Issues of interaction, presence, and performance in an online course. Journal of Asynchronous Learning Networks, 6, 21-40.

Pradana, M., Elisa, H. P., & Syarifuddin, S. (2023). Discussing ChatGPT in education: A literature review and bibliometric analysis. Cogent Education, 10(2), 2243134. https://doi.org/10.1080/2331186X.2023.2243134

Richardson, J. C., & Swan, K. (2003). Examining social presence in online courses in relation to students’ perceived learning and satisfaction, JALN, 7, 66–88.

Reddy, Y. M., & Andrade, H. (2010). A review of rubric use in higher education. Assessment & Evaluation in Higher Education, 35(4), 435–448. https://doi.org/10.1080/02602930902862859

Taylor, B., Kisby, F., & Reedy, A. (2024). Rubrics in higher education: An exploration of undergraduate students’ understanding and perspectives. Assessment & Evaluation in Higher Education, 49(6), 799–809. https://doi.org/10.1080/02602938.2023.2299330

Yuan, M., Zeng, J., Wang, A., & Shang, J. (2021). Would It Be Better if Instructors Technically Adjust Their Image or Voice in Online Courses? Impact of the Way of Instructor Presence on Online Learning. Frontiers in Psychology, 12. https://doi.org/10.3389/fpsyg.2021.746857