STAT1008 Study Guide - Midterm Guide: Binomial Distribution, Confidence Interval, Dependent And Independent Variables
Document Summary
When numbers are in categories they are categorical e. g. income. Sampling bias occurs when the method of selecting a sample causes the sample to differ from the population in some relevant way. If sampling bias exists, then we cannot trust generalizations from the sample to the population. Also if the population is likely to give an answer that is not truthful e. g. working out more than they do, not drinking (for people underage) What one value (or dot) in the distribution represents contest is non bias bell shaped. Se from the bootstrap distribution good approximation. Represents the statistic from a sample taken (without replacement) from the entire population. In both cases we compute the same statistic (mean, proportion, or whatever) and use the same sample size. Need to know the data values for the entire population. Represents the statistic from a sample taken with replacement from the original sample. Only need to know the values in one sample.
