You have finally begun to collect your data. You now have context for your study via the collected dataset. At this point, you'll probably have to review your preregistration and any other documents that indicate your methodology to confirm your analysis method makes sense. If they do, that's great; you can just ignore this blog post and move on. However, if it does not, then you are going to need to update your resources. So, let's take a brief moment to take a look at the update process.
Marking all Necessary Documents
In documents, I recommend updating the information in two locations. First, update the information within the text of the document. As this document may be released publicly or used within a report, it makes sense that the compiled information is up-to-date on what is currently done within the study. Second, provide an update block within your document or, if you have a version control system, in the commit history containing what change was made and why. This simply keeps track of how the study evolved with more context. Depending on your lab, sponsor, publisher, or other reviewer, you will need to keep a detailed log about how and why something changed after the initial methodology was approved. Additionally, by understanding the how and why something had to be changed, you now have more context for the future if you try to run a similar study without needing to change any pre-approved documents.
Updating a Preregistration
Most preregistration sites are version controlled, so a similar process is done to the previous section: change the information and provide an update message as to how and why. You might need to do this multiple times depending on how specific your preregistration is (e.g., if you put exact dates and a delay occurred).
Reporting
When reporting in a final document, which we will talk about more in a later post, you should be honest about whatever changes had occurred from your initial preregistration. I would limit this to a few sentences or a paragraph within a 'Discussion' section. This provides more transparency to any reader who might think of an idea you already tried or could not try because of some limitations on the compiled resources. If this is not included in the official document, add it as a supplementary material instead, which should be treated as a changelog of how your study evolved into the final reported one.
Some Additional Thoughts
After Data Analysis
After data analysis, updating existing documents becomes more complicated. This is because you already have context of the end result, so any additional changes would no longer be confirmatory in nature. As such, you would need to report the initial results from the data analysis along with any post-analysis, or exploratory, in a separate section. You should not replace any of the existing text in your methodology as it would not accurately reflect what the study was constructed for. Replacing the methodology would result in HARKing, which is highly discouraged within the research community.