Two ad agencies want to understand issues that are important to students. Agency A will interview 50 students whose names are randomly chosen from school enrollment data. Agency B will interview 50 randomly selected people from those who respond to a social media posting. Which method will produce a fair sample of the student population and why?