Using AI to Write README Files for Research Software

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-research-ai-research-software-readme-creators

1. Which of the following is something AI can typically generate for a research software README?

   1. A complete citation for the software even when no reference paper exists
   2. Installation instructions, usage examples, and citation information
   3. A fully tested and debugged version of the software
   4. Proof that the code produces statistically significant results

2. A researcher asks an AI to generate a citation for their newly published Python package, but no academic paper exists yet. What should the researcher expect?

   1. The citation section should be left completely blank in the README
   2. AI will create a fake DOI to make the citation look complete
   3. AI can only format a citation if the researcher provides the reference details
   4. AI will automatically find and cite any related publications

3. Before publishing a README that an AI has generated, what essential manual step must the researcher perform?

   1. Run every install and example command to verify they work
   2. Translate the README into all major languages
   3. Hire a professional editor to polish the prose
   4. Submit the README to a journal for peer review

4. In the context of research software, what does 'reproducibility' primarily refer to?

   1. The property that someone else can reproduce your exact results using your code and documentation
   2. The ability to run the same analysis on different operating systems
   3. The amount of memory the software requires to function
   4. The speed at which the software executes its computations

5. Why should a README for research software include reproducibility steps specifically?

   1. To meet journal requirements for open-source code
   2. To prevent others from modifying the original code
   3. To make the software run faster
   4. So that other researchers can independently verify and build upon the work

6. When generating a README with AI, which combination of elements would the AI be best equipped to handle?

   1. Debugging runtime errors and optimizing performance
   2. Formatting installation commands and documenting expected outputs
   3. Creating usage examples and explaining error messages
   4. Writing original code features and fixing bugs

7. What is the fundamental limitation preventing AI from generating accurate citations in all cases?

   1. READMEs do not typically include citations anyway
   2. AI refuses to work with bibliographic data
   3. AI does not understand academic formatting rules
   4. Citations require information that must come from external sources not present in the code

8. A researcher is preparing a README for their GitHub repository containing a new climate modeling tool. Which elements should definitely be included?

   1. The entire source code pasted into the README
   2. Only the author's biography and acknowledgments
   3. A list of all GitHub issues ever reported
   4. Purpose, dependencies, entry points, expected outputs, and citation information

9. Why is AI unable to test the research code itself when generating a README?

   1. Testing is unnecessary for research software
   2. Testing requires actually executing the code in a real environment with dependencies installed
   3. AI is legally prohibited from running code
   4. AI lacks the ability to read code files

10. In research software documentation, what does 'entry points' typically refer to?

    1. The main programming functions a developer would modify
    2. The location where the code is hosted online
    3. The commands or scripts a user runs to execute the software
    4. The inputs required to train a machine learning model

11. How are software dependencies related to reproducibility in research?

    1. They are unrelated to reproducibility
    2. Dependencies are only relevant for commercial software
    3. They must be precisely documented so others can recreate the exact computational environment
    4. They should always be the latest versions available

12. What is the risk of publishing a README without first testing the documented commands?

    1. Nothing—the AI guarantees all commands work
    2. The README will be rejected by version control systems
    3. The software might become more popular than intended
    4. The README might contain commands that look correct but fail when users try them

13. What is the primary purpose of including a README file in a research software repository?

    1. To increase the number of GitHub stars the repository receives
    2. To satisfy the requirements of the funding agency only
    3. To meet GitHub's aesthetic requirements
    4. To enable other researchers to understand, install, and use the software

14. Why is it important to document the 'expected outputs' of research software in a README?

    1. Expected outputs are required by copyright law
    2. So users can verify their run produced correct results and identify problems
    3. They make the software run faster
    4. Expected outputs determine the software's licensing terms

15. Which statement best describes what AI can and cannot do when creating a research software README?

    1. AI can test the code and verify citation accuracy
    2. AI can only generate the title of a README
    3. AI can generate documentation text but cannot verify the code works or create non-existent citations
    4. AI cannot help with README creation at all