Sampling and Types

The study of only some individuals in the population is called Sampling.

A sample is a subset of the population that represents the entire population.

- Types of Samples

Probability

Non-Probability

Non-Probability Sample

This research includes using available samples, using volunteer subjects, and purposive samples.

- Available Sample- This is a group of subjects that are readily and easily available for study. These may be people at a mall or a zoo. The issue with this is that there are unknown qualities of error. Some believe that they do not represent the population and therefore are not valid. This sample is usually used for questionnaires and pilot studies.

- Volunteer Sample- This form a non-probability sample since subjects are not mathematically chosen. Volunteer subjects differ from non-volunteer ones. According to Rosenthal and Rosnow they identified certain features of volunteer subjects compared to non-volunteer such as exhibit higher levels of education, occupational status, are more sociable. Volunteer sample should be used carefully because it also has unknown quality of errors.

- Purposive Sample- These subjects are chosen because of the certain characteristics they posses. This is usual used in advertisement research where the subjects are chosen on the basis of what product they use.

Probability Sample

In probability sample every person in the population has an equal chance of being a subject. It is random sampling without replacement.

There is a use of random numbers to generate random samples. For example if one want to analyze 10 prime time television programs out of a 100 programs to determine how elderly people are portrayed. The 100 programs will be numbered and then there will be random selection of numbers from 0-99.

Sampling using telephone survey is done using Random Digit Dialing (RDD). The problem with this is that many numbers are found to have been disconnected. So if a sample of 100 is needed, it is best that a sample of 300 be created. Random number generation is possible using a variety of method, however two rules have to be valid.

Ø Every subject in the population must have an equal chance to be selected.

Ø Selection process must be free from subjective intervention of the researcher.

In Systematic Random Sampling every ninth subject is selected from the population. However the accuracy of this method depends on the sampling frame or the complete list of members in the population. Telephone directories are inadequate sampling frame because they do not list all numbers. Also another issue with Systematic Random Sampling is periodicity, where the arrangement of the order of items in population list could be biased.

Simple Random Sample is preferred choice.

Stratified Sample is the approach used to get adequate representation of a sub-sample. Stratified sampling can be applied in two ways.

o Proportionate Stratified Sampling which includes with sizes based on proportions of population. If 30% of population is adults ages 18-24, then 30% of total sample will be subjects of this age group.

o

Cluster Sampling uses the process of selecting samples in groups. One can divided the state into districts, zips codes and make groups of subjects.

Snowballing Sampling also know as referrals. A few qualified subjects are contacted, and through them names relatives and friends are acquired who may also be qualified to study.