MODR 1730 Study Guide - Quiz Guide: Stratified Sampling, Inductive Reasoning, Sample Size Determination

115 views3 pages

Document Summary

Random sample: every item in populaion has equal chance of appearing. Representaiveness is when sample is similar to populaion in many relevant ways. Random sample is most likely representaive, but hard to achieve that: problems of sampling: size and representaiveness. Sample size depends on variability of populaion; less variable means you"ll have beter generalizaions and require smaller sample. Any study that has a sample of college students and target populaion of adults generally has unrepresentaive sample. Can"t test drugs only on males and expect them to work on women too: improving a sample. Representaiveness is the best to aim at, randomness is impossible. If doing a quesionnaire, need to consider who is going to get the quesionnaire and also who is going to return it: straiied sampling. Construct sample based on same composiion of populaion, more representaive. Can use this in cases where reliable background knowledge of regulariies is known: guidelines for evaluaing inducive generalizaions.