When I worked with a university partnered research lab, it wasn't uncommon to have a heavily biased sample based on the university student population. The stude

When I worked with a university partnered research lab, it wasn't uncommon to have a heavily biased sample based on the university student population. The students were significantly easier to tap into and often cheaper if compensation was involved. With that said, if there's an IRB process for any sort of human subject work, it's always skirting along the ethically grey area to do this. So, not only is your data biased towards a specific subpopulation (concentrated age groups, often concentrated socioeconomic statuses, etc.), faculty are often on the verge of unethical conduct with students. 98% is pretty high even for the cheapest of faculty I interacted with.