Chapter 7 Sampling Flashcards
generalizability and its other term
ensuring that the sample is a good representation of the pop. of interest; representativeness
sampling error
extent to which characteristics of individuals selected for the sample differ from those of the population
main component of a biased sample
not all members of a pop. have an equal prob. of being included
which validity do biased samples lack?
external
main component of an unbiased sample
all members of the pop. of interest have an equal prob of being included
probability sampling vs non-prob sampling
prob: every member of pop of interest has an EQUAL AND KNOWN chance of being selected for the sample, random
non-prob: produces biased samples, not random
5 techniques of random sampling
SRS
systematic
cluster
stratified
oversampling
how is SRS characterized?
equality: each member of pop of interest has an equal chance of being chosen
independence: selecting 1 person doesnt affect the likelihood of selecting any other
disadv of SRS
time-consuming and can be difficult
systematic sampling
randomly generate a number N and select every Nth person/thing
cluster sampling
pop of interest is divided into clusters (usually naturally occurring groups ex: location)
multistage sampling
variation of cluster sampling: subset of a subset of clusters
advantages of cluster
easier to contact participants and dont need a sample frame for entire population
sampling frame
list of all members of a population
stratified random sampling
divides the pop into strata/subgroups that are of interest in hopes of including participants from distinct strata
3 steps of stratified random sampling
- divide pop into strata
- randomly select cases from each subgroup
- add up all participants from dif strata together
sample sizes in stratified random sampling can be of 2 types:
equal or proportionate
oversampling
variation of stratified sampling; one or more groups is overrepresented
when is oversampling most appropriate?
when the distribution of subgroups isnt really equal
random sampling
using a random method to create a sample so each member of the pop of interest has an equal chance of being selected
random assignment
used in experiments to assign participants to groups at random
random sampling increases which validity?
external
random assignment increases which validity?
internal
random assignment helps to eliminate _______
confounds
when can nonprob/ non random sampling be used?
when external validity is not vital to the study
4 techniques of non random sampling
convenience
purposive
snowball
quota
in non prob sampling what does the researcher know about a participant’s chances of being selected?
nothing
convenience sampling
selecting participants who are readily available
purposive sampling
researchers judge and select who they think the best participants will be
snowball sampling
participants are asked to recommend others to participate
quota sampling
select specified proportions of certain kinds of participants
does sample size affect representativeness?
No
what is a typical statistically valid sample size?
1000-2000 cases
what quantifies sampling error?
margin of error of the estimate
an increase in sample size is shown to be related to a decrease in what?
margin of error