A schools principal wants to determine which type of speaker the students prefer to invite to an assembly for the entire student population which surgery method would produce the best representative sample?

For a survey like this, you want to take a random sample. It would be a good idea to use an alphabetical list of the students. Then, you could randomly select students from the list.

The principal would want to avoid surveying all the same type of students. For example, don't just ask the athletes.