Participant IDs almost match. What should you fix before merging?
A conservative way to handle spaces, zero-padding, duplicates, and typo-like IDs without turning every near match into the same participant.
Clean formatting; review identity
Removing a trailing space from P006 is not the same decision as saying P008X and P008 are the same person. The first is formatting. The second is an identity judgment. Keep those two kinds of fixes separate.
Zero-padding is easy to spot, but still worth documenting
If one system exports P0004 and another uses P004, the difference may be harmless. It is still useful to record the cleanup so a collaborator can understand why those rows were joined.
A collision is a stop sign
If a cleanup would turn two different source IDs into the same final ID, do not apply it automatically. Go back to the source files and determine what happened. A normalization that collapses two participants is no longer a formatting fix.
Duplicates and source-only IDs need different treatment
A duplicate such as P010 appearing twice in one source is a within-file problem. An ID such as P999 appearing only in a follow-up file is a source-only record. Both deserve review, but neither should quietly change the participant list.
A small example
P0004 → P004 likely formatting difference P006 → P006 whitespace cleanup P008X → P008 possible near match; review P010 appears twice duplicate P999 only in source report, do not silently add
What RDM does with these cases
RDM surfaces cleanup candidates, duplicates, source-only IDs, and conservative near-ID suggestions for review. It blocks cleanup that would create an ID collision instead of silently forcing a match.