Sampling & Sampling Distributions

Probability and probability distributions are a major part of statistics because from them, people can make inferences about a population using sample information.

An important sample statistic is the sample mean. The data is normally collected using either probiblistic or non-probabilistic procedures and terminology may vary depending on the design sample.

Sampling errors occur if the sample is not chosen to closely match the sample data with the actual data in the population. Measurement, data collection and human errors can distort the data.

Probiblistic data uses random or systematically chosen data. Every "n" has an equal chance of being chosen from "N".

Non-probilistic data is chosen when something is known about the data. Stratified or cluster sampling for example. Quota and convenience sampling are also forms of non-probiblistic sampling.

Having regard for samples and their probability distributions, there is a theory to describe the center known as the central limit theory. It states that regardless of the shape of a population used to calculate the  mean of sample, as the sample size gets larger, the shape of the probability distribution will approach a normal curve.

In calculation of a normal curve, it is important to know the standard deviation of a sample. If "s" is not known, the one would calculate "t" instead of "z" using the t-distribution.

If the population size is known you can do a sample population correction. Standard error * Square root of (N - n / N-1).