SAMPLING

• Sampling – it is the process of deciding who the subjects will be and then selecting them.
o Selecting subjects is an important part of any research regardless of its design, and it is a particularly
critical issue in survey research.
• The population consists of all people, animals, or objects that have at least one characteristic in common—for
example, all undergraduates form one population, etc. It is almost never possible to study the entire population;
instead, we rely on samples.
• A sample of subjects - is a group that is a subset of the population of interest.
o Data collected from samples can be used to draw inferences about a population without examining all its
members.
• How accurately we can generalize our findings from a given sample to a population depends on its
representativeness, or how closely the sample mirrors the larger population—more precisely, how closely the
sample responses we observe and measure reflect those we would obtain if we could sample the entire population.

PROBABILITY SAMPLING
• Probability sampling involves selecting subjects in such a way that the odds of their being in the study are
known or can be calculated.
• We begin by defining the population we want to study.
o For example, our target population might be all women born in 1975 and now living in Seattle,
Washington. (It would take time, but we could count them and establish each woman’s odds of being in
the sample.)
• A second condition for probability sampling is that the researcher must use an unbiased method for selecting
subjects, such as flipping a coin, drawing a number out of a hat, or using a table of random numbers.
o This process is called random selection, meaning that any member of the population has an equal
opportunity to be selected.
o Random selection is also a common assumption of the statistical tests used most often to analyze data.
SIMPLE RANDOM SAMPLING
• Simple random sampling is the most basic form of probability sampling
• In this sampling method, a portion of the whole population is selected in an unbiased way.
• Through random sampling, we can find out what the population is like without studying everyone.
• To obtain a simple random sample, all members of the population being studied must have an equal chance of
being selected.
SYSTEMATIC RANDOM SAMPLING
• It is an elegant variation of random sampling
• In cases where all members of a population are known and can be listed in an unbiased way, a researcher may
select every nth person from the population.
O The n is determined by the size of the population and the desired sample size.
STRATIFIED RANDOM SAMPLING
• Done when the population is known to contain distinct subgroups.
• A stratified random sample is obtained by randomly sampling from people in each subgroup in the same
proportions as they exist in the population.
 • By using stratified random sampling to include management, you ensure that their views are fairly represented in
the data
• There are two advantages to stratified sampling:
o First, the subsets are sampled separately so that important minorities or subgroups are represented in the
total.
o Second, different procedures can be used with different subgroups to maximize the usefulness of the data.
• This strategy is desirable when the various subgroups are likely to respond differently.
CLUSTER SAMPLING
• When the population of interest is very large, it is often too costly or impractical to randomly select subjects one
by one.
• Cluster sampling - Researchers sample entire clusters, or naturally occurring groups, that exist within the
population.
• As with all forms of probability sampling, participants are randomly selected, but whole groups of people are
selected rather than individuals.
• The main advantage of cluster sampling is that the researcher can sample data efficiently from relatively few
locations.
• A potential disadvantage of this approach is that subjects within clusters may resemble one another.
O The people in one county, for instance, might be quite similar in economic status, education, ethnicity,
and even age. For that reason, it is desirable to sample many clusters and to obtain as large a sample as
possible.

NONPROBABILITY SAMPLING
• In nonprobability sampling, the subjects are not chosen at random.
QUOTA SAMPLING
• In quota sampling, researchers select samples through predetermined quotas that are intended to reflect the
makeup of the population.
• Samples can reflect the proportions of important subgroups, but the individuals are not selected at random.
O For example, a newspaper might want to document campus attitudes toward nuclear arms. A reporter is
sent to a university with instructions to interview 40 students, half male, half female because the student
body is roughly half men, half women. There are no constraints on how the reporter selects people to
interview as long as the quota is filled.
• Such samples have human interest value and are sometimes used in public opinion polls, but quota sampling lacks
the rigor required in scientific research and is low in external validity.
CONVENIENCE SAMPLING
• Convenience sampling is obtained by using any groups who happen to be available
• This is considered a weak form of sampling because the researcher exercises no control over the
representativeness of the sample.
• Despite this drawback, convenience sampling (also called accidental sampling) is probably done more often than
any other kind.
• It is convenient, it is certainly less expensive than sampling the whole population at random, and it is usually
faster.
• It might not be valid to generalize observations beyond the group studied.
PURPOSIVE SAMPLING
 • Purposive sampling – Done when nonrandom samples are selected because the individuals reflect a specific
purpose of the study
• As with other forms of nonprobability sampling, researchers must be extremely careful about generalizing the
results obtained using a purposive sample to any other members of the population.
SNOWBALL SAMPLING
• In snowball sampling, another form of nonprobability sampling, a researcher locates one or a few people who fit
the sample criterion and asks these people to locate or lead them to additional individuals.
• Snowball sampling is used predominantly for sampling very small, uncommon, or unique populations at times
when researchers do not know who the population members are or how to contact them.
O Suppose you wanted to survey a sample of animal rights activists. After locating one activist who agrees
to be part of the research, you might ask this individual to help you find other activists and then ask them
to help find others, and so on.