Probability and Non-Probability Samples

Probability and Non-probability Samples

Probability Samples means that each unit in the sample has the same possibility to be chosen in the investigation. It is widely used in the sample method. For example if the teacher wants to know the smoking rate of the students in the class, every student in the class has the same opportunity to be chose. I think this much more fair for every unit in the sample, so the result will be much more accurate. Probability samples, even ones without especially high response rates, yielded quite accurate results. (Yeager, Krosnick, Chang, Layitz, Levendusky, Simpser,& Wang, 2011). Probability sample fits in the small size sample. Non-probability samples means that the investor choose sample in their own convenience and in nonrandom manner. Although it is also used but I think non probability samples have many problems. It is not as accurate as probability sample, because the sample investor chose was by his convenience and not every sample has the same opportunity. Non-probability sample is difficult to predict when such inaccuracy will occur, and because probability samples manifested consistently high accuracy, researchers interested in making accurate measurements can continue to rely on probability sampling with confidence. (Yeager, Krosnick, Chang, Layitz, Levendusky, Simpser,& Wang, 2011)

The principles of the probability sample are：larger sample size has the less sampling error, but it will cost much more money. So it’s not a good way to design a large sample size, it depends on the economic conditions. Investor can collect the information from the people in different age and different level; to estimate the sampling error. The results of the survey can be used to infer the overall. For example, in a survey if 5% of the respondents are given a certain answer, so investigators can percentage by combining sampling error, conveyed to the overall situation. In another hand probability sample also cost much money than...