PSYC 200W Lecture Notes - Lecture 8: Simple Random Sample, Nonprobability Sampling, Cluster Sampling
Sample - subset used to collect data from
Sampling - process by which a researcher selects a sample of participants for a study
Representative sample - a sample from which one can draw accurate, unbiased estimates of the
characteristics of a larger population
Sampling error - characteristics of individuals selected for sample are slightly different from
characteristics of general population
Error of estimation/margin of error - degree to which data obtained from sample deviates from the
population. Factors that affect error of estimation -
Sample size: larger the sample, the better it represents the population and leads to a smaller
margin of error
Population size: if the sample has 200 people and total population is 400, the error of estimation
is smaller
Variance of data: greater variability in the data makes it difficult to estimate the population
accurately and gives a wider margin of error
Probability sampling - probability of choosing sample is known - used for descriptive research
Simple random sample: Every possible sample of the desired size has the same chance of being picked.
Requires a sampling frame (list of population) to select randomly from
Systematic sampling: taking every nth individual for the sample
Stratified random sampling: divide the population into strata that share the same characteristics and
then randomly selected
There is a drawback to stratified and random sampling - we don't always have a sampling frame
Cluster sampling: there are clusters that naturally occur in a population and then random clusters are
chosen and everybody in that cluster is used in sample
Multistage cluster: sample larger clusters, then smaller clusters within the large until we obtain
the final sample
Advantages - don't need sampling frame because we have clusters, each cluster is based on
geographical proximity and thus requires lesser effort and time
Nonresponse problem - failure to obtain responses form individuals who are selected for a sample
Lack of time
Illness
Literacy or language problems
Disinterest
Solve it - give incentives, follow up, assess differences between responders and nonresponders
Misgeneralization - results can be misleading if researcher generalizes results to a population that is
different from the one the sample is drawn from
Nonprobability sampling - don't know the probability of a case being chosen
Convenience: participants that are readily available
oIf examining relations between variables/hypothesis testing, then convenience is okay
oMust carry out replication to extrapolate it to other groups/samples
Quota: convenience sample where certain kinds of participants are obtained in particular
proportions
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Sample - subset used to collect data from. Sampling - process by which a researcher selects a sample of participants for a study. Representative sample - a sample from which one can draw accurate, unbiased estimates of the characteristics of a larger population. Sampling error - characteristics of individuals selected for sample are slightly different from characteristics of general population. Error of estimation/margin of error - degree to which data obtained from sample deviates from the population. Sample size: larger the sample, the better it represents the population and leads to a smaller margin of error. Population size: if the sample has 200 people and total population is 400, the error of estimation is smaller. Variance of data: greater variability in the data makes it difficult to estimate the population accurately and gives a wider margin of error. Probability sampling - probability of choosing sample is known - used for descriptive research.