•
o Equal probability of selection designs (EPS), in which each element of
the population has an equal chance of being included in the sample. This
uniformity makes EPS surveys relatively simple to interpret. Forms of
EPS include Simple random sampling (SRS) and systematic sampling.
•
o Probability-proportional-to-size designs (PPS), in which 'larger'
elements (according to some known measure of size) have a higher chance
of selection. This approach is common in business surveys where the
object is to determine sector totals (e.g. "total employment in
manufacturing sectors"); compared to EPS, concentrating on larger
elements may produce better accuracy for the same cost/sample size.
•
o Stratified random sampling approach, in which the population is
divided into subpopulations (called strata) and random samples are then
drawn separately from each of these strata, using any probability sampling
method (sometimes including further sub-stratification). This may be done
to provide better control over the sample size (and hence, accuracy) within
each subpopulation; when the variable/s of interest are correlated with
subpopulation, it can also improve overall accuracy. Another use for
stratification is when different subpopulations require different sampling
methods - for instance, a business survey might use EPS for businesses
whose 'size' is not known and PPS elsewhere.
Random sampling:
In a probability sample (also called "scientific" or "random" sample) each member of the
target population has a known and non-zero probability of inclusion in the sample.[4] A
survey based on a probability sample can in theory produce statistical measurements of
the target population that are:
• unbiased, the expected value of the sample mean is equal to the population mean
E(ȳ)=μ, and[5]
• have a measurable sampling error, which can be expressed as a confidence
interval, or margin of error.[6]
Within probability sampling there are specialized techniques such as stratified sampling
and cluster sampling that improve the precision or efficiency of the sampling process
without altering the fundamental principals of probability sampling.
Non-random sampling:
Many surveys are not based on a probability samples, but rather by finding a suitable
collection of respondents to complete the survey. Some common examples of non-
probability sampling are[10]:
• Judgement Samples: A researcher decides which population members to include
in the sample based on his or her judgement. The researcher may provide some
alternative justification for the representativeness of the sample.
• Snowball Samples: Often used when a target population is rare, members of the
target population recruit other members of the population for the survey.
• Quota Samples: The sample is designed to include a designated number of people
with certain specified charateristics. For example, 100 coffee drinkers. This type
of sampling is common in non-probability market research surveys.
• Convenience Samples: The sample is composed of whatever persons can be most
easily accessed to fill out the survey.
In non-probability samples the relationship between the target population and the survey
sample is immeasurable and potential bias is unknowable. Sophisticated users of non-
probability survey samples tend to view the survey as an experimental condition, rather
than a tool for population measurement, and examine the results for internally consistent
relationships.