Page 30 - Read Online

Koss et al. Art Int Surg. 2025;5:116-25 https://dx.doi.org/10.20517/ais.2024.91 Page 118

METHODS
Study design and setting
This cross-sectional study received an exemption from the institutional review board at the University of
California, San Francisco. The primary aim was to understand how patients seeking GAS utilize ChatGPT.
An anonymous 5-minute online survey was developed using Prolific (London, UK, 2024), an online
research platform, and distributed to eligible users in January 2024. The full survey, including all questions,
is provided in Supplementary Materials. Prolific offers advantages as a research survey platform, including
access to a diverse participant pool, built-in demographic screening tools, and a straightforward process for
targeting specific populations. The platform is frequently used in academic research for its ability to deliver
reliable data quickly and cost-effectively.

The survey included multiple-choice, free-response, Likert scale, and checkbox-style questions. Participants
were compensated $1.10 for completing the survey, in accordance with Prolific’s recommended rate of
$12/h, ensuring fair and ethical compensation. We acknowledge that participant compensation may
influence responses, as individuals may feel incentivized to complete surveys quickly rather than
thoughtfully. To mitigate this, we designed the survey to be concise, clear, and straightforward, minimizing
opportunities for rushed or disengaged responses. Additionally, Prolific’s payment structure is designed to
provide fair compensation without incentivizing low-quality data. We chose Prolific due to its transparency,
demographic screening capabilities, and robust reputation for high-quality data in academic research.
Alternative platforms, such as MTurk, have been criticized for issues such as lack of demographic diversity
and unreliable responses. Traditional methods of recruitment, such as in-person surveys or phone
interviews, were not feasible for this study, given time and resource constraints.

Study population
Individuals were included if they were English-speaking, aged 18 and older, based in the United States, and
indicated that their current gender differed from their assigned gender at birth. Participants were selected
based on pre-screening criteria available within Prolific, such as age range and self-reported interest in
health-related topics. All participants had pre-noted demographic data from when they signed up for a
Prolific account which allowed them to be included in the study. These criteria were established to ensure
that participants were likely to be representative of individuals who might seek information online about
gender surgery. On Prolific, participants provide demographic information when creating an account,
allowing us to identify eligible individuals based on their self-reported demographics. To ensure participants
were truthful about their demographic information, we utilized Prolific’s pre-screening tools, which require
users to provide verified demographic details before they can participate in studies. Additionally, we
included attention-check questions and survey logic consistency checks to detect and exclude participants
providing potentially unreliable responses. Efforts were made to obtain a racially and ethnically diverse
sample by distributing the survey to 200 Prolific users, approximately half of whom identified as White and
half as non-White.

Data collection
Data were collected through the online survey, which gathered information on demographics, interest in
GAS, and interactions with ChatGPT. To ensure data quality, multiple attention checks were embedded
throughout the survey, and participants who failed these checks were excluded from the final analysis. In
the survey, GAS was defined as any of the following procedures: mastectomy, breast augmentation, facial
feminization, facial masculinization, vaginoplasty, phalloplasty, and/or metoidioplasty. For participants who
used ChatGPT to access information about GAS, additional questions explored the scope of information
sought and how this information compared to other resources.