Meet the Methods series: Measuring Research Participants' Sexes and Genders
Issue 6 - November 2025

1. Intersex people exist (those people born with a variation in their sex characteristics), most of whom were assigned a binary sex. If you want to ascertain whether your sample includes both intersex and endosex people (those who are not identified as having a difference in their sex development), you will want to ask not only about sex assignment, but also whether participants were born with a variation, difference, or divergence in their physical sex characteristics.^{Footnote 3}
2. Acknowledge discomfort, as some participants including some trans participants may find this question distressing and harmful – be prepared to offer why you are asking and how the data will be used. This can be done in writing or verbally in advance of the question being asked. For example, "this question will be used to determine which subsequent sections of the survey you will see," where those sections may only be relevant to people assigned a particular sex (i.e., questions about pregnancy).
3. Be clear on what isn't being measured, since sex assignment at birth cannot be reliably used to make claims about legal sex, experiential sex, or current anatomy or physiology. All you know from these data is someone's external genital configurations at birth.

1. Amend for jurisdictional specificity, since legal sex markers will need to be appropriate and representative of the geographic location where the study is taking place (e.g., M, F and X in Canada, for adults).
2. Consider whether the intention of changing legal sex is worth measuring, as some participants in your study may wish they could change their legal marker, and this may be fruitful data.
3. Be clear on what isn't being measured, since legal sex cannot be reliably used to make claims about sex assignment, experiential sex, or current anatomy or physiology. All you know from these data are whether the person has an M, F, X or other marker on their government-issued identity documents.

Do you have [insert body part]?

To the best of your knowledge, are you capable of [insert physiological function]?

1. The specific question asked will need to be precise and free of proxies. For example, having a uterus does not necessarily mean that someone is capable of pregnancy; having a penis does not necessarily mean that someone produces sperm.
2. Explain the purpose as some participants may be surprised or even offended to be asked these questions. Be prepared to offer why you are asking and how the data will be used. This can be done in writing or verbally in advance of the question being asked. For example, "We are asking about whether you have a cervix, so that you will be asked subsequent questions about cervical cancer screening in this survey."
3. Gender and facets of sex are entangled. For example, even if you ascertained someone's hormone levels through blood tests, behaviour affects hormone levels and vice versa. It is impossible to separate sexed variables from the influence of the social environment.
4. Determine whether the anatomy you are asking about needs to have been present at birth, or whether surgically created anatomy are also being asked about. For example, one person might have a vagina that was present at birth, and another may have a vagina that was constructed via vaginoplasty; one might have a penis that was present at birth, and another might have a penis that was constructed via phalloplasty. You may have good faith reasons for wanting to know this information, or for limiting eligibility for a project to only those with anatomy present at birth or surgically created anatomy.

1. Balance inclusive options with data utility, where determining which and how many response options to include will depend on how you intend to use these data. The list may also include cultural and/or regional terminology. A wider range of response options may allow participants to more accurately self-identify but may result in having insufficient statistical power to run analyses. Consider offering two gender identity questions – one where participants can freely describe their gender(s) with nuance and complexity, and a second with fewer response options, where you make clear that they are being asked how they'd like to be categorized for the purposes of stratification or comparison.
2. Consider the implications of "select one" versus "select all that apply." For example, "select one" may lend itself to more straightforward analyses, but does not leave room for people who have more than one gender identity. If "select all that apply" is used, consider in advance how you will attend to the complexity of responses – will you overwrite their choices? Will you group each person in each stratum that they selected? Will you differentiate between participants who selected only one gender identity, from those who selected that same gender identity, alongside one or more additional options?
3. Consider the implications of open-ended write-in boxes, as they have the benefit of allowing participants to freely describe themselves. However, they may also require care and expertise when coding or recoding, depending on the complexity of those responses.
4. Remember that Two-Spirit is not analogous with many of the Western gender identity labels, and ought to be ascertained after having determined that a participant is Indigenous. Two-Spirit can be used to refer to facets of gender and/or sexuality, but it's more than that, for many. Check out the IGH Meet the Methods on Two-Spirit identities for more.

Are you a person of trans experience? Are you a person with trans history? – or other single measures that ask whether someone is trans outright.

Or

Asking about and then comparing reported sex assignment and current gender identity for concordance or discordance (or branchedness/coincidence^{Footnote 6}) – called the two-step method.

Or

3x3 method – where participants tell you their gender(s) and place themselves into a 3x3 grid, representing cis, trans, and allogender categories, and then binary, nonbinary and allobinary categories. They then narrate the reasoning behind their selections.

1. Avoid othering, which happens when we add 'trans woman' and 'trans man' as gender identity response options, alongside 'woman' and 'man.' This suggests that there are 'wo/men' and then 'trans wo/men' as separate categories. If 'trans woman' and 'trans man' are listed, they should appear alongside 'cis woman' and 'cis man' – with a note that this question is therefore ascertaining gender identity and gender modality simultaneously. Remember that not everyone claims 'cis' or 'trans' as aspects of their gender identities.
2. When using the two-step method, consider how you will categorize participants who claim multiple gender identities, both concordant and discordant (or branched and coincident). For example, a participant may indicate being assigned female, and that they identify as both a man and nonbinary. Is this person to be categorized as trans, cis or allogender?^{Footnote 7}
3. When using the two-step method, consider the ordering of questions, as research has found that asking gender identity first, and sex assignment second is preferable to many trans participants. The reverse puts sex assignment into a place of primacy within a survey.^{Footnote 8}

1. Do not assume gender identity and expression will match. Without asking explicitly, you cannot reasonably assume that all women in your sample are also feminine, and where you therefore theorize about how societal perceptions of femininity impacted the outcomes in your data. Some people's gender expressions will align with what is typically expected (e.g., a man who is masculine), but some people's gender expressions are not neat, linear, expressions of their 'corresponding' identities.
2. Self-perception matters but may not always map onto how others perceive any given person. Consider whether you need to be additionally asking your participants how they think others see, perceive and experience them. If a nonbinary person perceives themselves as androgynous but anticipates that the world is receiving them as a masculine (cis or trans) woman - this misrecognition in terms of both gender identity, modality and expression may be useful in explaining some of how this person is treated and thus, their experiences in the world.