Suppose you have developed a miracle diet plan. You experiment using 100 individuals and discover that they lost their weight. Do you comfortably say that this diet will be effective for everyone in the world? Probably not. This is where the ideas of comparing random sampling vs random assignment are introduced. According to a report published by the NIH, less than 1 percent of clinical trials are representative of the entire population. More so, 95 percent of psychological research is based on college students.