Statistical sampling refers to the study of populations by gathering information about and analyzing it. It is the base for a great deal of information, ranging from estimates of average height in a nation to studies on the impact of marketing to children. Numerous professions use statistical sampling, including psychology, demography, and anthropology. Like any study method, however, this method is prone to errors, and it is important to analyze the methods used to conduct a study before accepting the results.
This process begins with a definition of the population the scientist wants to study, and the variable which he or she wants to measure. For example, someone might want to know the average weight of elementary school children. Next, the scientist decides how to collect the desired data. In the previous example, the scientist might travel to schools with a scale, send questionnaires out to doctors or parents, or try to access school health records. Many researchers try to measure directly, rather than relying on selfresponses, because this way the results are consistent.
Once the population, variable being measured, and method have been defined, the scientist decides how to accurately sample the population so that the collected data is representative of a larger group. In other words, statistical sampling does not involve measuring the desired variable in every individual of the population being studied; a selection of individuals is used to generalize results. Generally, the larger the sample size, the better the results.
The most common system is random sampling, in which a scientist generates a list of random individuals from a central database. Some scientists use cluster sampling, in which a population is divided into a bunch of small clusters and each cluster is studied extensively. Others might use systematic sampling, in which every nth person in the population is studied. The most dangerous and unreliable selection system for statistical sampling is convenience sampling; someone standing on a street corner with surveys is using convenience sampling, which can yield highly inaccurate results.
After the data is collected, the researcher analyzes it and uses it to make generalizations about a population. In studies which rely on statistical sampling, the method used is usually clearly detailed, so that other scientists can decide whether or not the method was valid. An invalid method can cause sampling error, which would call the results of the study into question.
parmnparsley Post 3 
@ GenevaMech The difference in statistical vs nonstatistical sampling is human bias. Statistical sampling is any sampling that is done by random sampling methods and uses probability theory to measure the sample risk and evaluate the results of the sample. 


Glasshouse Post 2 
@ GenevaMech A population is just what it sounds like. It is the entire population of something. A sample is a part of a population that someone chooses to analyze. In many cases, it would be impractical to collect data from or about an entire population so someone would take a sample. There are different ways to determine sample populations in statistics, but they should be representative of the larger population. Probability sampling uses a random device to determine the population that will be sampled to eliminate human bias. Cluster sampling can be used to determine a sample from a geographically scattered sample. Stratified sampling separates a population into groups and then takes representative samples of each group. Finally, multistage sampling uses multiple methods to get a sample that represents a large, complex population. 
GenevaMech Post 1 
What is the difference between population and sample in statistics? Also, what is the difference between statistical and nonstatistical sampling? Someone please help me out. 