POL S 15 Lecture Notes - Lecture 12: Statistical Inference, Sampling Error, Statistic
Document Summary
Five 4 different "good" pollsters the same polling data in september 2016 and ask them to estimate the poll results themselves using this data . All pollsters came to different predictions given the same data. Indicative of how hard it is to get at an undefined population. How you weight the sample trying to make samples look more represented (counting individuals for more to increase the presence/value of those underrepresented) Some group of things (actors, people, countries) that we often cannot study (either practically or theoretically difficult), so all we can do is construct a sample drawn from the population. Tools of statistical inference help us draw conclusions about the hypothesis in the population based on the sample. Difference between the true proportion of the population and what you get from your sample = sampling error. Sampling error = the difference between the parameter (population) and statistic (sample) Occurs because our sample is not the whole population.