The key is the use of statistically derived random sampling procedures. These ensure that survey results can be defended as statistically representative of the population. Surveys that do not follow these procedures can produce results that lead to misguided market research, strategic, or policy decisions. Any so-called "survey" in which no attempt is made to randomly select respondents, such as call-in readers' or viewers' "polls", is likely to produce results that in no way reflect overall public opinion--even if many thousands of individuals participate. It is true that sampling randomly will eliminate systematic bias The mathematical theorems which justify most frequentist statistical procedures apply only to random samples. http://www.ma.utexas.edu/users/mks/statmistakes/RandomSampleImportance.html no author,
COMMON MISTEAKS MISTAKES IN USING STATISTICS: Spotting and Avoiding Them 4/10/12
Moore and McCabe (2006), Introduction to the Practice of Statistics, Third edition, p 219 Sample Distribution and Sampling Error
Distributions of populations of scores have been discussed. However, a single score does not accurately represent the population. A sample, or subset, of the population is a better estimator of the population. Alternatively, a sample that has received some treatment can be compared to the original population. Just as there is a distribution of scores for a population, there is a distribution of samples (or sampling distribution) for a population. These distributions become perfectly normal when all scores, or all samples, are included. All possible scores/samples are rarely available or possible, but the population can be reasonably represented by proper sampling.******** If the samples are randomly chosen, they are likely to be more representative of the population than a single score is. In addition, the larger the random sample is, the better the representation will be. The sample will likely be different than the population, but the...
Please join StudyMode to read the full document