You want to determine if eating more fruit reduces a person's chance of developing cancer. You watch people over the years and ask them to tell you how many servings of fruit they eat each day. You then record who develops cancer. Is this a cross-sectional, retrospective, or prospective study?