Citing Research Software Properly: From Stata to PyTorch to That Custom Pipeline

End-of-lesson check

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

1. A graduate student submits a paper to a journal that now requires software citation. What is the main argument for why this requirement improves scientific practice?

   1. It increases the number of citations received by software developers
   2. It allows other researchers to locate and use the exact same software version that produced the published results
   3. It ensures that all software used in research becomes open source
   4. It prevents researchers from using proprietary analysis tools

2. Which set of elements constitutes the core FORCE11 software citation principles that journals and funders now expect?

   1. Creators, title, version number, publisher, and identifier
   2. Algorithm description, computational complexity, test dataset, and validation metrics
   3. License type, download count, GitHub stars, maintainers, and commit history
   4. Authors, abstract, keywords, DOI, and publication date

3. A researcher wants to cite a Python package hosted on GitHub that has not been uploaded to Zenodo or any other DOI-issuing repository. What can an AI tool realistically help with in this situation?

   1. Create a new DOI for the repository
   2. Assign a persistent identifier that works like a DOI
   3. Generate a citation using the GitHub URL and commit hash as the identifier
   4. Register the software with the appropriate domain authority

4. What is the fundamental limitation of using an AI tool to generate software citations for a project that was completed six months ago?

   1. The AI lacks access to the internet to look up the software
   2. The AI cannot distinguish between stable and beta versions
   3. The AI cannot retroactively capture the exact package versions that were installed at the time of analysis
   4. The AI will always generate incorrect author names

5. A computing environment statement for a reproducibility appendix typically includes which of the following pieces of information?

   1. Operating system, programming language version, and key library versions
   2. The funding sources that supported the research
   3. A detailed explanation of the algorithms implemented
   4. The names of all co-authors who ran the code

6. Why might two researchers using the same software package produce different results, even when using the same version number?

   1. One researcher is using a pirated copy of the software
   2. The software version numbers are not reliable indicators of functionality
   3. Software versions are assigned randomly by developers
   4. Different random seeds or underlying data changes can produce divergent outputs even with identical software versions

7. A researcher deposits their custom analysis script in Zenodo and receives a DOI. Later, they modify the script and want to cite the updated version. What is the correct approach?

   1. Create a citation without any DOI since the software has been modified
   2. Continue citing the original DOI since it always points to the latest version
   3. Use the original DOI but add a note explaining the changes
   4. Deposit the modified version as a new version in Zenodo, which provides a new DOI that specifically identifies this updated release

8. Which statement accurately describes what AI tools can contribute to software citation workflow?

   1. AI can decide which journal is most appropriate for submitting the research
   2. AI can generate properly formatted citations and recommend appropriate identifiers based on available metadata
   3. AI can determine whether the research findings are statistically significant
   4. AI can automatically write and execute the entire analysis code

9. A PhD student uses conda to create an isolated environment with specific package versions for their thesis analysis. What problem does this practice solve?

   1. It allows the student to avoid documenting which packages were used
   2. It automatically publishes the code as open source
   3. It makes the analysis run faster on supercomputers
   4. It ensures that dependencies are locked at the time of analysis rather than relying on potentially changed package versions later

10. When preparing a reproducibility appendix that lists every software dependency, which piece of information is least critical to include?

    1. The exact version number of each dependency
    2. The programming language and version used
    3. The funding grant number that supported the development of each package
    4. The name of each software package used

11. A researcher cites a software package using its Zenodo DOI. What property does this identifier provide that a simple URL does not?

    1. Persistence — the DOI remains valid even if the software moves to a different hosting location
    2. Transparency — the DOI reveals the full names of everyone who ever contributed code
    3. Immunity — the DOI prevents others from forking or modifying the software
    4. Priority — the DOI establishes who published the software first

12. Why do discipline-specific citation styles sometimes require different formats for software compared to the FORCE11 general principles?

    1. Different fields have different conventions for how detailed software contributions should be acknowledged
    2. There is no variation across disciplines in how software should be cited
    3. Software citation styles are determined exclusively by the software developers
    4. FORCE11 principles are only applicable to biological sciences

13. A research team downloads a custom machine learning pipeline from a colleague's private GitHub repository (not public) and uses it in their study. How should they cite this software?

    1. They should not cite software they did not write themselves
    2. Cite it as a personal communication or unpublished software, noting the source and any available version information
    3. GitHub repositories cannot be cited since they are not peer reviewed
    4. The team should wait until the software is published in a journal before citing it

14. What is the primary purpose of a code and data availability statement in a research paper?

    1. To prove that the researchers followed all applicable ethics guidelines
    2. To list all authors who contributed to the project
    3. To satisfy journal requirements that have no impact on reproducibility
    4. To inform readers where they can access the exact data and software used in the study

15. A student is writing a paper and asks an AI to generate citations for all the Python packages they used. The AI produces citations with DOI numbers. What should the student verify before including these citations in their paper?

    1. That the DOI numbers actually resolve to the correct software and version
    2. That the AI has permission to generate citations for commercial software
    3. That the packages are listed in alphabetical order
    4. That the citations use their professor's preferred font style