With sampling the population, I think a poor example of sampling would be to use a bias sampling by using a surveyor to go on campus and ask all 1000 students which movie they would prefer. This would take the surveyor a long time and that would not be a sample, it would be the entire population. The surveyor would get good answers, but it would definitely take too long since it isn't a sample. With sampling the population, I think a better example would be to take about 350 students as a sample of the 1000 and organize the results in a histogram. You can take a bar chart and put the 350 students on there with their genders and figure out which movie preferences the students liked by gender and grade and get your results. The relationship of a sample to a population is just a smaller percentage of the entire value. The population is the whole value. I would classify my samples as convenience for the poor example and cluster for the better example. explain why cluster sampling would not work here