--- tags: Superprompting_guide --- # ChatGPT Examples ## Example Sets for Teaching, Research, and Administration ## Teaching Examples 1. Objective: Enhance students' critical thinking skills. Prompt: Provide a list of five open-ended questions related to John Stuart Mill's utilitarianism that encourage students to think critically about the theory. Include potential ethical dilemmas and real-world applications. Limit your answer to approximately 150 words. 2. Objective: Teach students how to write a literature review. Prompt: Provide a step-by-step guide on how to write a literature review for a research paper in the field of social sciences. Discuss the process of selecting relevant sources, organising the content, and synthesising the information. Include at least two reputable sources on literature review best practices. Limit your answer to approximately 250 words. ## Research Examples 1. Objective: Develop a research proposal for studying the psychological effects of social media usage. Prompt: Draft a research proposal exploring the psychological effects of social media usage on young adults aged 18-24. Provide a brief background, research question, methodology, and expected outcomes. Discuss any potential ethical concerns and how they can be addressed. Include at least three reputable sources to support your proposal. Limit your answer to approximately 500 words. 2. Objective: Evaluate the effectiveness of different data collection methods for a marketing research project. Prompt: Compare and contrast the advantages and disadvantages of three data collection methods (surveys, interviews, and focus groups) for a marketing research project investigating consumer preferences for sustainable products. Discuss the appropriateness and limitations of each method for this specific research context. Include at least two reputable sources to support your analysis. Limit your answer to approximately 300 words. ## Administration Examples 1. Objective: Develop a rubric for assessing students' oral presentations. Prompt: Create a rubric with five criteria for evaluating students' oral presentations in a university-level communications course. For each criterion, provide a brief description and a three-level scale (e.g., poor, satisfactory, excellent) with corresponding point values. Include at least two reputable sources on assessment best practices. Limit your answer to approximately 300 words. 2. Objective: Optimise the organisation of a module syllabus. Prompt: Provide recommendations for organising a module syllabus in a clear and logical manner. Discuss the essential components to include, such as learning objectives, assessment methods, course delivery plan, and assessment policies. Offer suggestions for improving readability and accessibility. Include at least two reputable sources on syllabus design best practices. Limit your answer to approximately 200 words. --- # Conclusion: Differences between GPT3.5 and GPT4 GPT4 is a more advanced model compared to its predecessor, GPT3.5. When using GPT4, you can expect: 1. **Improved understanding**: GPT4's ability to understand complex and nuanced prompts has been enhanced, providing more accurate and relevant responses. 2. **Better context retention**: GPT4 can better retain context throughout longer responses, resulting in more coherent and consistent answers. At the moment, ChatGPT4 uses the '8k' token model which means that it can produce up to the equivalent of ~ 5 A4 pages of text in one go. 3. **Increased knowledge base**: GPT4's training data is more extensive and up-to-date, offering more comprehensive and accurate information in its responses. 4. **Refined language generation**: GPT4 is better at generating fluent, natural-sounding text, leading to higher-quality responses. It still seems more like American English in most outputs. Try adding an addition prompt term such as: >Ensure your outputs are written using spelling, terminology and idioms suitable for a UK English audience* Despite this, GPT4 may still produce incorrect or nonsensical answers occasionally. Remember though that we're just at the start of the LLM AI journey- it will only improve from here on. As a user, it is essential to critically evaluate the generated content and use your expertise to assess its reliability and relevance.