Surveying Adolescents During a Pandemic: Comparison of Adolescents Recruited via Social Media vs. Schools

School-based surveys of adolescents can be logistically difficult and exclude students who do not attend school. Social media recruitment could be a promising strategy to recruit representative samples of adolescents. However, few studies have compared adolescent survey data collected via different methodologies. Our team was conducting a school-based survey when the COVID-19 pandemic closed all schools, necessitating a shift to online adolescent recruitment. To achieve our goal of obtaining a sample of high school students throughout California, we placed ads on social media. We compared the adolescents recruited in schools with those recruited on social media on demographic characteristics, mental health, and substance use. The sample of students recruited in schools (N = 737) and adolescents recruited via social media (N = 953) did not differ significantly on gender or substance use. However, compared with school-based recruitment, social media recruitment yielded a higher proportion of boys, whites, and Asians and a lower proportion of girls, Hispanic/Latinx adolescents, and those who spoke other languages at home. The social media sample had significantly higher levels of depression and anxiety symptoms and perceived stress than the school-based sample. Results indicate that social media can be useful for recruiting adolescents for survey research, especially if strategies such as Spanish-language social media ads are used to recruit and consent Hispanic/Latinx adolescents and those with non-English-speaking parents. This method could potentially replace school-based surveys in cases where schools are unwilling to participate in research, or it could be used to supplement school-based samples. Advantages and disadvantages of both methods are discussed.