When you are asking generative AI, such as ChatGPT or Google Gemini, to help you research a problem you need to formulate a prompt. The term "prompt" refers to the input you enter into the AI's search field: the question you are asking AI to answer, or instructions you are giving AI indicating what you are looking for.
To get the best output from AI, it's important to create a good prompt. The more detail you can give the AI about what you are looking for, the better the results you will receive.
Follow the CLEAR framework for crafting your prompts. The following method and quotations are from Lo (2023), p. 2, cited below:
- (C)oncise: "brevity and clarity in prompts"
- "A concise prompt removes superfluous information, allowing AI language models to focus on the most important aspects of the task, resulting in more pertinent and precise responses";
- (L)ogical: "structured and coherent prompts"
- "To create a logical prompt, ensure that the provided information follows a natural progression and that the relationships between concepts are evident";
- (E)xplicit: "clear output specifications"
- "Explicit prompts provide precise instructions regarding the desired output format, content, or scope, thereby reducing the likelihood of receiving unanticipated or irrelevant responses from the AI model" ;
- (A)daptive: "flexibility and customization in prompts"
- "Adaptability entails experimenting with various prompt formulations, phrasings, and temperature settings in order to establish a balance between creativity and concentration"; and
- (R)eflective: "continuous evaluation and improvement of prompts"
- "Adopting a reflective perspective enables users to evaluate the performance of their AI model based on user feedback and their own assessments, identifying areas for improvement and adjusting their approach accordingly".
For more information about the CLEAR method, see:
