Sampling Flashcards
what are the two ideals in quantitative research work?
generalizability and objectivity
what do generalizability and objectivity in quantitative research rely on?
quality data
why is selecting participants well important for quantitative research?
if you select participants badly, you’re going to end up with false conclusions
why is objectivity in sampling for quantitative research important?
objectivity in sampling is essential to produce accurate, reliable, and valid research findings that can be generalized to the larger population
why is generalizability important in sampling for quantitative research?
if you don’t sample correctly, you can’t generalize your findings to a larger population
how was sampling used in South Africa when standardizing international measures?
South African citizens, those living in SA for a long time, etc need to be familiar with South African culture
South Africa is not a homogenous population. Differences in:
Legacy of apartheid
Race
Religions
Traditional cultures
Languages
Education
Need a large enough sample which has distributions across all of these characteristics to then draw conclusions about the entire SA population
what should you look for when sampling?
generally want to sample so that our sample matches the wider population on some key demographic variables
what does generalizability in sampling mean?
obtaining a sample that represents the population
what is the first step when sampling for generalizability?
starting with a sampling frame
what is key when sampling for generalizability?
make sure the sample is representative
what is a sampling frame?
the area within which you’re going to sample - thinking of key demographics, traits of the people that you want to sample
what are the second and third steps to sampling for generalizability?
choosing sample well, and then choosing sampling method
what does a sampling frame do?
gives you the idea of the greater population and gets you kind of this huge list of people who could potentially be in your study
how is random selection different to random assignment?
While random selection involves choosing participants randomly from your population, random assignment is when you randomly distribute participants into the intervention or control group.
what sampling method is best for strong external validity?
random selection. This will involve methods like simple random sampling, stratified random sampling, or cluster sampling, depending on your study design.
what does random selection sampling create in a study?
strong external validity
what does strong external validity mean?
results can be applied to the wider population
what does random selection and random assignment control for?
ensure that the sample is not biased
what are the selection methods for probability sampling?
sample selected at random
what are the selection methods for non-probability sampling?
sample selected based on the subjective judgement of the researcher
what are the chances of selection for probability sampling?
equal for all
what are the chances of selection for non-probability sampling?
varies, not guaranteed
what is the level of generalizability for probability sampling?
high
what is the level of generalizability for non-probability sampling?
limited (sample does not accurately represent the population)
what are the types of probability sampling?
simple random sampling (random selection), systematic sampling, stratified sampling, cluster sampling
what are the types of non-probability sampling?
convenience, purposive, snowball
what is simple random sampling?
random selection - every unit in the sample frame has an equal chance of being selected (probability sampling)
what is systematic sampling?
choosing units at fixed intervals (every nth unit) (probability sampling)
what is stratified sampling?
identify subgroups within the population, determine their proportions, and then randomly select a proportional number of units from each subgroup (probability sampling)
what is cluster sampling?
divide the population into clusters, sample clusters randomly, sample randomly from each cluster (probability sampling)
what is convenience sampling?
anyone who volunteers to participate (non-probability sampling)
what is purposive sampling?
selection of people who are typical of the population (non-probability sampling)
what is snowball sampling?
starting with some members of the population and then asking for their help to identify other possible participants (non-probability sampling)
what is quota sampling?
starting with subgroups based on desired characteristics, setting quota, then non-random selection (non-probability sampling)
what is sampling error?
statistical error that occurs when one does not select a sample that represents the entire population of data
what does sampling error aim to do when inferring something about a population?
Be able to quantify how far off our sample statistic could be from the population statistic (sampling error)
Have some assurance that the sample is representative of the population (i.e., minimize the chance of bias)
how can the prevalence of sampling error be reduced?
by increasing sample side
what are the categories/types of sampling errors?
population-specific error, selection error, sample frame error, non-response error
what is the population-specific sampling error?
error occurs when the researcher defines the target population incorrectly from the beginning. They might be investigating the wrong group entirely, leading to irrelevant findings.
what is the selection sampling error?
error arises when the sampling method itself favors certain members of the population over others. This can happen in non-probability sampling methods or even in probability sampling if not implemented carefully.
what is the sample frame sampling error?
error occurs when the sampling frame, the source from which the sample is selected, is inaccurate or incomplete. This occurs before a sample is selected. If the frame doesn’t represent the entire population well, the chosen sample won’t either
what is the non-response sampling error?
error happens when potential participants selected for the sample do not respond to the survey, interview, or experiment. If non-respondents differ systematically from the respondents, it can bias the results
