The premise
Pre-registration thoroughness catches the analytic decisions researchers should make in advance; AI helps draft comprehensively.
What AI does well here
- Generate pre-registration drafts following AsPredicted, OSF, or registry-specific templates
- Surface pre-specified decisions researchers might forget (alpha levels, exclusion criteria, multiple comparisons)
- Compare draft pre-registration against existing exemplars in the field
- Generate the deviations log for inevitable post-hoc decisions
What AI cannot do
- Substitute for substantive methodological thinking
- Eliminate the discipline of actually following pre-registration
- Replace peer review of pre-registration
End-of-lesson check
15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-research-AI-pre-registration-creators
What is the primary purpose of pre-registering a research study before data collection?
- To prevent researchers from making analytic decisions after seeing results
- To guarantee that findings will be statistically significant
- To obtain funding approval from institutional review boards
- To automatically publish results in academic journals
Which of the following is an explicit output when using AI to draft a pre-registration?
- A dataset containing the study results
- An automated email to ethics committee approval
- A final peer-reviewed manuscript ready for submission
- An AsPredicted or OSF-formatted pre-registration document
Which of these research decisions would typically be included in a pre-specified decisions checklist?
- The conclusion the researcher wants the data to support
- The final sample size after data collection is complete
- The specific alpha level and exclusion criteria to apply
- The names of co-authors who will review the paper
What is the purpose of a deviations log in the context of pre-registration?
- To track participant recruitment timelines
- To list funding sources that supported the research
- To document any analytic decisions that differ from the original pre-registered plan
- To record changes made to the pre-registration document after submission
A researcher uses AI to draft a pre-registration. What crucial limitation should they remember?
- AI can substitute for the researcher's own methodological reasoning
- AI cannot replace peer review of the pre-registration
- AI eliminates the need for the researcher to follow their pre-registered plan
- AI will guarantee their study receives ethics approval
How does pre-registration contribute to research transparency?
- By ensuring all research questions are answered correctly
- By eliminating the need for informed consent procedures
- By automatically correcting statistical errors in submitted manuscripts
- By making the planned analysis publicly visible before data collection
Which platforms were mentioned as examples where pre-registrations can be submitted?
- TikTok and Instagram
- PubMed and Google Scholar
- AsPredicted and OSF
- YouTube and Vimeo
Even with AI assistance in drafting a pre-registration, what must researchers still do themselves?
- Request that AI make all final methodological decisions
- Avoid documenting any deviations from the original plan
- Submit the pre-registration without reviewing its accuracy
- Follow the pre-registered plan when conducting analyses
What does the term researcher degrees of freedom refer to?
- The number of participants required for statistical power
- The physical workspace where researchers conduct experiments
- The choices researchers can make about data analysis after seeing results
- The methods used to select research assistants
What type of comparison can AI perform for a drafted pre-registration?
- Comparison to the researcher's personal social media history
- Comparison to unrelated commercial advertising campaigns
- Comparison to existing pre-registration exemplars in the same field
- Comparison to fictional study designs from novels
A researcher inputs their study design into an AI system. What should they expect the AI to produce?
- A pre-registration draft and analysis script outline matching the plan
- A published article ready for journal submission
- A guarantee that the study will be replication
- A complete statistical analysis with final results
What is included in the pre-specified decisions checklist that AI can generate?
- Personal information about the research team
- The specific p-value needed for publication
- Alpha levels, exclusion criteria, and multiple comparisons corrections
- The journal where results will be submitted
What does OSF stand for in the context of research transparency?
- Open Science Framework
- Operational Study Format
- Original Scientific Findings
- Online Statistical Forum
When researchers make post-hoc decisions that differ from their pre-registration, what should they do?
- Only inform their close colleagues
- Document them in a deviations log
- Pretend the original plan was always followed
- Hide these decisions to avoid criticism
What information must a researcher provide as input for AI to draft a pre-registration?
- The final statistical results and conclusions
- Research questions, hypotheses, methodology, and analysis plan
- The complete raw dataset
- A list of potential reviewers for the manuscript
