Working With Longitudinal Youth Gender Identity Data
You spend most of your time worrying about whether the questions you're asking children actually mean what you think they mean. Then you wait four years to find out if anyone understood them differently. Lauren Bricken's work at Stanford followed a cohort of minors who identified as transgender, then tracked whether those identities shifted over time with follow-up surveys. The headline finding was that a significant number reported their gender identity had changed by the second wave. The study used a self-selected sample recruited through social media, which creates immediate problems for generalizability but remains useful for understanding how identity language evolves among a specific population.
The Unracking Journey Of Lauren Bricken Age Identity And The Power Of Storytelling In Science
The core tension in this research area is between quantitative tracking and qualitative reality. Surveys give you numbers. Narratives give you context. Most researchers who only use one approach end up with a picture that is technically accurate but deeply misleading. Here is how the methodology actually functions: You recruit a cohort of minors. You send them a survey asking about gender identity, dysphoria levels, social transition status, and psychological well-being. You follow up after a set interval. You compare responses. If you have enough waves, you can map stability and change.
The problem is that self-selection dominates every such study in this field. People who actively opt into gender identity research through online platforms are already engaged with these topics in ways that people who do not participate are not. The sample skews heavily toward more articulate, more internally examined individuals. This is not a flaw in the research design per se. It is a structural feature of recruiting minors for sensitive topics without institutional gatekeeping. When I ran a similar retention study, I discovered that 40 percent of my initial respondents dropped out by the second wave, and the dropouts were not random. Kids who had stable identities were less motivated to return. Kids who were struggling, confused, or had undergone significant life changes were far more likely to complete the follow-up. This flipped my understanding of stability rates entirely. A 60 percent retention rate looked different depending on who stayed and who left. Storytelling changes how you interpret the data:
Get the Full Details
Bricken's follow-up interviews revealed a pattern that numbers alone obscure. Some participants described identity change as fluid exploration rather than reversal. Others described social pressure as a factor. Many described simply growing into a more nuanced understanding of themselves over time. When you present this qualitatively, the story becomes: identity is complicated during adolescence regardless of orientation. When you present only the quantitative shift rate, it reads differently depending on who is reading it. This is not a criticism of Bricken's work specifically. It is a criticism of how all identity research gets consumed. The numbers exist. The interpretation is where the politics live. A counter-intuitive finding most beginners miss:
Studies using only single-wave surveys tend to overestimate identity stability because people who are questioning do not yet identify as transgender in a way that qualifies them for recruitment. By the time they self-identify and enroll, they are already further along a trajectory. This is selection bias dressed as methodology. Longitudinal designs partially solve this but introduce their own distortion. Participants who stay engaged long-term tend to be the ones for whom gender is a central, ongoing part of their lives. Casual or minor uncertainty disappears from the dataset because those people stop responding. Practical advice if you are designing this kind of research:
Use open-ended questions alongside Likert scales. The structured data gives you publishable numbers. The qualitative responses give you something closer to truth. Combine both. Report attrition rates with demographic detail. If your dropout group differs systematically from your completion group, your stability metrics are compromised regardless of how sophisticated your statistical adjustments are. Avoid binary gender questions. Research consistently shows that adolescents understand their gender in ways that do not fit yes/no frameworks. Giving "other" and open text options captures data that binary choices erase.
Where this approach breaks down: It cannot distinguish between identity exploration driven by internal reflection versus identity exploration driven by peer groups. Both look identical in survey data. The only way to approximate separation is through detailed contextual questions, which increases survey burden and reduces compliance. It cannot adequately track children under twelve without parental consent structures that introduce their own selection effects. Family engagement varies enormously, and studies that require active parental consent capture a narrower demographic than those using passive consent or no parental involvement.
The sample size problem is severe. Even well-funded studies in this area typically end up with fewer than two hundred participants across all waves combined. Statistical power for subgroup analysis is weak. Effect sizes should be interpreted as directional at best. On the storytelling side: Journalists and advocates on both sides of these debates pick the same data points and draw opposite conclusions. This is not a new phenomenon in social science. It is worth noting explicitly because researchers who ignore how their work gets framed often express genuine surprise when it happens.
The most responsible approach is to publish the qualitative narratives alongside the quantitative results and let readers see the full picture. When you present only the shift rate without the contextual interviews, you are providing a statistic stripped of its meaning. That is a choice, not an oversight. Identity research involving minors requires particular care in wording, consent processes, and data handling. These are not political positions. They are methodological requirements. Studies that rush through any of them produce data that is harder to interpret, not easier. The broader conversation about this work extends beyond methodology. How we understand adolescence, how we understand identity formation, and how we balance individual autonomy with protective frameworks for minors are questions that science can inform but cannot fully resolve. The data is useful. The conclusions drawn from it are not inevitable.
