Surveys are supposed to turn complicated human experiences into useful data. Unfortunately, some questionnaires approach sexual orientation and gender identity with the subtlety of a vending machine: pick one button, accept whatever drops out, and please do not ask why the options make no sense.

For sexual- and gender-diverse teens, poorly designed questions can mean being misclassified, grouped into an inaccurate category, or excluded from the results altogether. A nonbinary student may be forced to choose “male” or “female.” A pansexual teen may select “bisexual” because it is the closest available option. A questioning student may abandon the survey after realizing there is no honest answer.

These are not merely awkward questionnaire moments. When surveys overlook LGBTQ+ adolescents, public health agencies, schools, researchers, and community organizations receive an incomplete picture of youth health. That can affect decisions about mental health services, bullying prevention, housing support, sexual health education, crisis intervention, and school safety.

The good news is that better measurement is possible. Researchers increasingly understand how inclusive wording, thoughtful survey design, strong privacy protections, and direct involvement from young people can produce more accurate data without turning a five-minute questionnaire into a graduate seminar.

Why Sexual- and Gender-Diverse Teens Disappear From Surveys

Many national and local surveys were originally built around assumptions that every respondent was either male or female, straight or gay, and completely certain about both. Adolescence, however, is a period when attraction, identity, language, relationships, and self-understanding may still be developing.

A teen may know who they are attracted to but not use a specific identity label. Another may identify as bisexual without having dated anyone. A transgender boy may describe himself as straight, while a nonbinary student may use several terms for sexual orientation depending on the context. None of these experiences fits neatly into a single old-fashioned checkbox.

The National Academies of Sciences, Engineering, and Medicine has emphasized that sex, gender identity, and sexual orientation are distinct concepts. Treating them as interchangeable creates measurement errors before the first spreadsheet has even been opened.

Sexual orientation has more than one dimension

Sexual orientation may involve identity, attraction, and behavior. Those dimensions can overlap, but they are not identical.

A student might identify as straight while reporting attraction to more than one gender. Another might identify as lesbian but have no history of sexual activity. A bisexual teen may currently be dating someone of a different gender, yet that relationship does not make the teen heterosexual.

Surveys that ask only about sexual behavior can overlook teens who have not been sexually active. Surveys that ask only about identity can miss young people who are questioning, who do not recognize the available labels, or who prefer not to adopt one. Researchers should select questions based on what they genuinely need to understand rather than pretending one item measures the entire universe of human attraction.

Gender cannot be measured accurately with one binary question

A question such as “Are you male or female?” may be interpreted as asking about sex assigned at birth, current gender identity, legal documents, physical characteristics, or what the respondent thinks the researcher wants to hear.

This ambiguity is especially consequential for transgender and nonbinary teens. A transgender girl might choose “female” because she is a girl. A researcher might later classify every student who selected “female” as cisgender. The student answered honestly, but the survey interpreted her answer incorrectly.

A widely recommended alternative is a two-step approach. One question asks about sex assigned at birth, while another asks about current gender identity and permits more than two responses. Researchers can then identify cisgender, transgender, nonbinary, and other gender-diverse participants more accurately.

The wording matters. In a study involving students from Pittsburgh public high schools, a broader two-step measure identified substantially more gender-diverse students than earlier surveys that had relied on a single “Are you transgender?” question. The study was local rather than nationally representative, so its percentage should not be applied to every American high school. Its central lesson, however, is difficult to ignore: change the question and the apparent population may change dramatically.

Most Teens Are Willing to Answer Inclusive Questions

One common objection is that questions about sexual orientation and gender identity are too confusing, private, or sensitive for adolescents. Research does not support the assumption that young people will simply refuse to answer.

A large U.S. study of adolescents and young adults led by researchers affiliated with Boston Children’s Hospital examined missing responses across questions about sexual orientation, gender identity, gender expression, sex assigned at birth, and changes in sexual orientation. About 95.7% of participants completed every relevant item. Only a small minority skipped multiple questions.

That does not mean every question was equally successful. Items about sex assigned at birth and past sexual fluidity were more likely to be left unanswered. The pattern suggests that researchers should improve wording and context, not throw the entire subject into the questionnaire recycling bin.

Child Trends has also tested sexual orientation and gender identity measures with middle and high school students. Its work supports using language young people understand, offering appropriate response options, and testing questions with racially, culturally, and developmentally diverse groups before launching a survey at scale.

How Narrow Survey Design Produces Bad Data

Limited response options

A survey may offer “straight,” “gay or lesbian,” and “bisexual,” followed by “not sure.” That is better than asking nothing, but it can still leave out pansexual, queer, asexual, demisexual, and other identities. It may also treat “questioning” as confusion rather than a meaningful description of a respondent’s current experience.

Researchers do not need to include a dictionary-sized list of every term ever used online. They can provide several common choices, allow respondents to select more than one, include a write-in option, and periodically update the measure with youth input.

Mutually exclusive categories

Some teens use multiple labels. A participant might identify as queer and bisexual, transgender and nonbinary, or asexual and panromantic. A radio-button format that permits only one selection forces the student to surrender part of the answer.

“Select all that apply” is often more accurate, although it creates additional work during analysis. That is not a design failure. It is reality politely informing the research team that human beings are more complicated than a drop-down menu.

Faulty skip patterns

Survey software frequently decides which questions appear next. If a student answers “no” to “Are you transgender?” the software might hide every later question about gender identity. That can exclude nonbinary students who do not personally use the word transgender.

Similarly, a survey may ask sexual health questions only of students who report an opposite-gender partner. Same-gender relationships, gender-diverse partners, and teens who use different language for their experiences may disappear from the analysis.

Proxy reporting

Some household surveys ask one adult to answer for everyone in the home. A parent may not know, understand, or feel comfortable reporting a teen’s sexual orientation or gender identity. The teen may not have disclosed it.

For personal identity measures, direct and confidential self-reporting is usually more informative than asking another household member to guess. Proxy reporting can be useful for household characteristics, but it is a shaky foundation for understanding a young person’s inner life.

Combining everyone into one LGBTQ+ category

Researchers sometimes combine lesbian, gay, bisexual, queer, transgender, nonbinary, and questioning participants into a single group because individual samples are small. Aggregation may improve statistical power, but it can hide major differences.

A bisexual cisgender girl may encounter different risks than a transgender boy. A Black nonbinary teen may experience racism and gender-based harassment simultaneously. A rural gay student may face barriers that are less common in a large city. Reporting one average for the entire LGBTQ+ population can produce a tidy chart and a very untidy understanding of reality.

Why Missing Data Matters

Survey data influences where money, staff, and attention go. If gender-diverse students are undercounted, a school district may underestimate the need for counselor training, inclusive anti-bullying policies, private changing areas, or support groups.

The CDC’s 2023 national Youth Risk Behavior Survey marked an important development by asking high school students about transgender identity at the national level. Its findings showed that transgender and gender-questioning students experienced higher levels of violence, poor mental health, housing instability, and school safety concerns than cisgender peers.

For example, approximately one-quarter of transgender and questioning students reported missing school because they felt unsafe, compared with fewer than one in ten cisgender male students. Numbers like these help schools understand that attendance problems may be connected to safety rather than laziness, bad alarms, or an unexplained national allergy to first-period algebra.

At the same time, the 2023 survey had a historically low overall response rate, driven largely by school participation. Even well-written questions cannot represent students who never receive the questionnaire. Inclusive measurement therefore requires attention to wording, sampling, school recruitment, response rates, weighting, and transparent reporting of limitations.

The Williams Institute has produced national estimates indicating that hundreds of thousands of U.S. teens identify as transgender. Other studies generate different percentages depending on their samples, definitions, response choices, and survey methods. Those variations do not prove that one group of teens suddenly appeared or vanished. They demonstrate why methodology must be explained whenever an estimate is reported.

What Better Youth Surveys Should Do

Ask separate questions about separate concepts

Sex assigned at birth, current gender identity, sexual orientation identity, attraction, behavior, and gender expression should not be blended into one mystery question. A survey does not need to measure every dimension, but it should clearly identify which dimension it is measuring.

Allow multiple and self-described responses

Inclusive options should cover common identities while allowing respondents to select multiple answers or describe themselves in their own words. Write-in responses should be reviewed rather than automatically discarded as unusable “other” data.

Ask relevant questions of everyone

Researchers should avoid assuming a student’s orientation or gender based on appearance, name, previous answers, or relationship history. Asking standardized questions of all participants reduces selective misclassification.

Protect privacy

Teens are more likely to respond honestly when they understand who will see the answers and how the information will be protected. School surveys should not display sensitive responses where classmates, teachers, or family members can easily view them.

Researchers should collect only the information necessary for the project, limit access to identifiable records, explain confidentiality in age-appropriate language, and establish safeguards for small subgroups. Publishing a tiny demographic cell from one rural school could unintentionally identify an individual student even when no name appears.

Use youth advisory groups

Sexual- and gender-diverse teens should help design, test, and interpret the surveys intended to describe them. Youth advisors can spot confusing terminology, unsafe administration procedures, missing identities, and assumptions that experienced researchers may overlook.

This is not about asking teenagers to run the entire statistical department between chemistry homework and soccer practice. It is about recognizing that lived experience is a form of expertise.

Preserve consistency without freezing language forever

Standardization allows researchers to compare results across years. Updating questions improves relevance and accuracy. Good survey programs need both.

When wording changes, researchers should document the revision, test old and new versions when possible, and warn readers when trend comparisons may be affected. A sudden increase in a population estimate may reflect better measurement rather than a rapid demographic transformation.

Report uncertainty honestly

Reports should disclose missing responses, excluded participants, sample sizes, weighting procedures, question wording, collapsed categories, and limitations. Small groups should not be erased merely because their confidence intervals look less photogenic.

Inclusive Surveys Benefit Every Student

Better sexual orientation and gender identity questions do more than improve LGBTQ+ statistics. Clearer surveys reduce confusion for everyone. They distinguish identity from behavior, avoid forcing assumptions onto participants, and help researchers understand how health and education outcomes differ across groups.

Inclusive data can also reveal protective factors. Surveys should not focus exclusively on crisis, bullying, depression, or suicide risk. They can examine school belonging, supportive families, trusted adults, affirming health care, community connection, academic success, resilience, and joy.

The Trevor Project and GLSEN have repeatedly documented the importance of affirming environments, supportive adults, inclusive school policies, and respectful treatment. Measuring only harm can unintentionally portray LGBTQ+ teens as collections of risk factors rather than young people with talents, relationships, ambitions, jokes, and impressively specific opinions about music.

Conclusion

Sexual- and gender-diverse teens are not inherently impossible to survey. They are often overlooked because questionnaires rely on narrow choices, outdated assumptions, poor skip logic, weak privacy protections, or categories that were never tested with the people expected to use them.

More inclusive measurement will not eliminate every uncertainty. Identity language evolves, adolescents develop at different rates, and small samples create analytical challenges. Still, imperfect complexity is better than confidently publishing an inaccurate picture.

When researchers ask clear questions, protect confidentiality, permit nuanced answers, involve youth in design, and disclose methodological limits, surveys become more than collections of boxes. They become tools for seeing students who have been present all along.