This artifact of the data comes from the Williams report not wanting to stake firm ground on the question of whether or not nonbinary-identified people automatically get the "transgender" label as well. In trying to be accomodating, the researchers allowed cis (i.e, non-trans identifying) people to label themselves as nonbinary even though that goes against a lot of the social and political movement work. The result is a completely unusable and unscientific report where respondents are allowed to check as many identities as they like without any working definition of what the terms mean in reality.