PSYCH291 Lecture Notes - Standard Deviation, Sampling Error, Statistical Parameter
Document Summary
Example) methodology of the literary digests 1939 presidential poll. Poll conducted by mailing postcards to telephone/car owners, magazine subscribers. Prediction based on 2 million returned postcards, landon predicted to win 57% of popular vote. Sample biased towards wealthy republicans rather than average new dealers. Received 23% response rate (only enthusiastic people/ those seeking change) Population: groups of all individuals to whom researchers intend to generalize their results. Census: study includes all members of population as participants. Sample: subset of population of interest that researchers observe in their study (studying 5% of waterloo undergrads to infer characteristics of entire group) Population parameter: attribute of population whose value the researchers are trying to estimate (eg. waterloo undergrads" confidence level in employment prospects) Sample statistic: value of attribute that researchers observe in sample (eg. average level of confidence reported by individuals whom participated in survey) Differences between sample estimate and population parameter due to random differences between sample and population.