In: Math
Explain what a biased sample is and how random sampling reduces the potential of obtaining a sample that is biased. In inferential statistics, why is a biased sample a bad thing?
A biased sample is the sample in which observations are not selected by using random sampling and the probability of selection of each observation is not same. There is a bias during the selection of observations and sometimes purposefully selection of some observations creates a biased sample. The random sampling method reduces the bias in the sample because we select each observation randomly and there is a same probability for selection of each observation or item under study. There would be more different reasons for biased sample such as measurement errors, mistakes, non-response, etc. In inferential statistics, the biased sample is a bad thing because the results based on the biased sample are not useful for the further prediction. The biased sample produces biased estimates for the population parameter and we cannot infer properly about the population under study. Due to lack of unbiasedness we cannot generate unbiased estimators for the population parameters. So, it is important to use random sampling for minimizing the bias in the sample.