Understanding Sample Design in Marketing Research: From Parameters to Sampling Error

1    10/3/2023    Sample vs. census  When a population is counted, information is gathered from or about each person. An example  would be choosing a portion of the objects in a larger group.    Parameter: a population's attribute or measure We compute statistics from sample data in order to  estimate the population. Statistic: a feature or measure of a sample.    The discrepancy between results from a sample and those that would have been achieved had  data been collected from or about every member of the population is known as sampling error.    Sampling Frame: The set of population components from which a sample will be drawn; the set  may include institutions, people, places, or other types of units.  - Customer databases, member directories, and lists created by data compilers are some examples  of often used sample frames.     Sample Designs: - Quota (know the characteristics of the population and know how to structure  the sample) Nonprobability Samples: Convenience, Judgment, Snowball (go out and discover  "certain" person by being recommended to other like persons by people who are in your sample)  When it comes to recruiting, this type of sample in general can be biased at times.  - Probability Samples: Simple Random, Systematic, Stratified, and Cluster -- this is comparable  to the lottery in that anyone could be chosen for it.    as a convenience 

2    Sample of a judgment:    A cluster sample is when a population is divided into parts and a few subsets are selected at  random to be sampled.  When you divide a population into sections and take EACH subset of each sector, you are using  a stratified sample.