Yeah but “both” genders is precisely why I don’t think it’s an appropriate grouping. It’s exclusive of gender non conforming people whom, I’d imagine, make up a significant portion of the sample for gender neutral names. While it may not be reported in the source data, a non negligible portion of the sample population may have legally changed their name to a gender neutral name without legally changing their gender.

1. Its possibly not an accurate representation given that trans men may be reported as women and vice Verda.

2. It reduces gender to a binary which is not an appropriate framing when gender neutrality is a component of the question (as is the case here)

A better grouping might be name as assigned at birth vs legally adopted name or self reported gender including non binary identities. If binary gender is the only possible grouping in the source data I’d question the usefulness of aggregating by group.