Effective prompting
Formulating effective prompts is a skill. There are different ways of improving your prompting skills. One way is to experiment with different ways of formulating your prompt, perhaps based on example prompts. In this way, you may develop an understanding of why a certain prompt works better than another. Alternatively, you may prefer a more structured approach. Using established models or frameworks can help you craft prompts that yield more accurate and relevant responses. Below, you will find advice on how to prompt more effectively.
Quick tips
When formulating your prompt, consider the following:
- express yourself clearly, include context, and make clear requests.
- adapt your prompt to the AI tool you are using. It is especially important how you formulate your prompts when using chatbots that do not have a specific or explicit area of use. Here you will find examples of AI tools for different purposes.
- prompt iteratively, that is, ask follow-up questions to improve the output.
- avoid using generative AI tools for calculations. Chatbots are generally poor at handling numbers.
- do not share copyrighted material and/or sensitive information such as personal data.
Less effective prompts
"Improve my text."
"Rewrite my text."
Why are these prompts less effective?
- They are unspecific, which means that the tool may not generate the text suggestions you were looking for. For example, new information or references may be added. In addition, the suggested structure and style may differ greatly from what you had in mind.
- Since unspecific prompts can lead to both substantial and diverse changes in the text, it may be difficult to evaluate the suggestions.
More effective prompts
"I am writing an assignment for a university course in physiology. Can you give me examples of things I can revise in this draft and also explain why I should make these changes?"
"Can you give me examples of how I can make the structure of my assignment even clearer?"
"Can you help me make my text more academic so that it suits a university-level nursing assignment? Please also explain why I should make the suggested changes."
Why are these prompts more effective?
- These prompts provide a clear context for your text, which will probably give you more useful suggestions.
- You are asking for suggestions instead of rewrites. Therefore, you are forced to consider suggestions before making changes, which likely will allow you to learn more.
- You are also asking for explanations as to why different suggestions have been made. This will make it easier for you to make informed choices, and you also learn more.
Less effective prompts
Why are these prompts less effective?
- There is no context or purpose. AI tools cannot determine what you think is important.
- There is no information about what the summary should contain. AI tools cannot summarise; they can only shorten text.
More effective prompts
Why are these prompts more effective?
- It is clear what you want.
- You can verify the answers and deepen your understanding of the key concepts on your own.
- You will get a basic understanding of the text, which will make it easier for you to tackle the entire text and deepen your understanding of its content.
Less effective prompts
Why are these prompts less effective?
- The first consists of only a few words, not a full question. In these cases, AI tools will usually produce less helpful answers.
- Your question is biased. The phrasing assumes that melatonin could negatively affect dementia. The AI search tool may replicate that bias in its answers.
More effective prompts
Why is the prompt more effective?
- They are complete question, not just a few words.
- The question is objective and does not introduce bias.
Less effective prompts
Why are these prompts less effective?
- They lack detail about the desired format or purpose.
- They do not specify what kind of summary or quiz is needed.
- They assume the AI can determine what is important without context.
- AI tools do not "summarize" in the traditional sense. They shorten text based on patterns, not judgment.
More effective prompts
Why are these prompts more effective?
- They provide context, such as program, course, and learning objectives.
- They clarify purpose (for example, exam preparation).
- They specify the format and language of the output you need.
Traci and Create: two models to create effective prompts
Prompt models, or prompt mnemonics, are useful tools to help you understand how to build an effective prompt. Two of these models are called TRACI and CREATE. Try out both to see which one better suits the tool you are using and the task you want to solve.
Traci
The TRACI model offers a helpful framework to guide your prompt formulation. Each letter represents a key component:
- T – Task Clearly define the task you want the AI to perform. For example, “Summarise a research article” or “Generate patient education material.”
- R – Role Specify the role or perspective from which the content should be written. This could be a medical student, clinician, librarian, or researcher.
- A – Audience Indicate who the content is intended for, such as fellow students, patients, faculty, or journal readers.
- C – Create Provide instructions on the format or structure of the output. Do you want a bullet point list, formal report, infographic description, or a patient-friendly explanation?
- I – Intent Clarify the desired impact of the content. Do you want the text to be informative, persuasive, educational, or is the intent to simplify complex information?
Less effective prompts
- “Give me the most important points from this text.”
“Write a take-home exam on x.”
Neither of these prompts provide context. Picking out “the most important point” from a text depends on context (What is most important to whom?). Thus, the answer could be several different things. The prompts lack information about the sender, recipient, purpose, and effect.
More effective prompts according to the TRACI model
Example 1
Please give me ten study questions from this text. Focus on questions that require me to think and reflect, not mere yes or no questions (task). I am a student on the x course (role). The person who will answer the questions is a fellow student (audience). The language must be correct (create).
Example 2
Summarise the information in the following two sources (task). I am a specialist nurse in surgical care (role) writing for my university teacher (audience) and do not want bullet points (create). The text should be credible (intent).
Both these prompts provide good context. Since you have used the prompt model, it is clear what you want from the tool. The choice of words is precise. Of course, the answers must still be carefully assessed and verified for accuracy.
Create
Create is an acronym of the components of this prompt model: context, result, audience, explanation, tone, edit.
- C – Context Include relevant background information and specify who is making the request. For example, “I am a medical student preparing for a pharmacology exam.”
- R – Result Clearly state what you want the AI to produce. This could be a summary, a set of study questions, a patient information leaflet, or a draft for a research abstract.
- E – Explanation Provide guidance on how the content should be structured or presented. Should it be formal, concise, visual, or conversational?
- A – Audience Identify who the content is intended for, such as fellow students, patients, instructors, or journal reviewers. This helps tailor the language and depth of information.
- T – Tone Specify the desired tone of the output. For example, should it be professional, empathetic, instructional, or neutral?
- E – Edit Refine the response by reviewing and adjusting based on the previous components. You may ask follow-up questions or request revisions to better meet your needs.
More effective prompts according to the CREATE model
Example 1: I am a student on the X program (context) and I want an outline for an oral presentation on X (result) to my classmates (audience) in a factual style (tone). (Depending on the response you receive, you may refine your search further by specifying what you want to improve. An example of would be: “Add three examples of y.”)
Example 2: I am studying for x (context) and want three abbreviated texts from the following sources (result) to be read by an academic (audience) and written in formal language (tone). (Depending on the response you receive, you may refine your search further by specifying what you want to improve. An example would be: “Explain the concept of x.”)
Both these prompts provide good context. Since you followed the model, it is clear what you want from the tool. The choice of words is precise. Of course, the answers must still be carefully assessed and verified for accuracy.
Do you want more examples of effective prompts?
Browse different "prompt libraries", such as the one from Maastricht university below.
Learn more

Keep in mind!
You are always responsible for your own learning and what you produce in your studies.
Make sure you do so with academic integrity, that is, be transparent about how you use AI tools and do not use them more than is permitted for your course.
Do not share personal information, sensitive data or copyrighted material with the tools.
Om du vill att vi ska kontakta dig angående din feedback, var god ange dina kontaktuppgifter i formuläret nedan