Sampling for growth studies

The purpose of sampling is to take measurements on a representative portion of the population so that the whole population does not have to be measured. Each observation in the sample can be thought of as representing a certain number of population members. The reciprocal of the number in the population represented by an observation is the sampling proportion.

Sampling schemes

Most schemes for obtaining samples fall into one of three categories. Sampling can be done: (1) from a complete or nearly complete list; (2) from a set of people who go somewhere or do something; or (3) in two or more stages (Fowler, 1984). The first category implies that there is a list of the population available in advance from which then to take a sample. Examples of the second category are sampling children who get services from a hospital or who shop at a store with a parent. The third category includes schemes such as sampling, in turn, states or provinces, counties or departments, neighbourhoods or villages, households, and then individuals.

Characteristics to evaluate sampling

Three characteristics important for evaluating a sampling scheme are: (1) comprehensiveness; (2) known probability of (or equal) selection; and (3) efficiency (Fowler, 1984). Comprehensiveness refers to whether everyone in the population of interest from which the sample is to be drawn had a chance to be selected into the sample. That is, a sampling scheme is not comprehensive if some people in the population are excluded from being possibly sampled.