Research & Statistics Flashcards
has the adv. of being quick and easy to do
doesn’t use a census
draw ppl in a haphazard & convenient way
non-probability sampling
involves a census
the adv of ___ is that if done correctly it should resemble the population
probability sampling methods
where every element (ppl in population) has an equal chance of being chosen
simple probability
instead of random sampling, you are choosing every nth element
n = the population size divided by the desired sample size
systematic
(probability sampling)
when you are interested in groups
1) divide the census in groups of interest
2) draw from each group using simple probably sampling
stratified (groups)
(probability sampling)
when elements are in ___(s)
1) randomly sample the ___
2) take rosters from each sample & then randomly sample elements from each cluster using simple probability
cluster
(probability sampling)
convenience, accidental
person on the street approach
haphazard
(non-probability sampling)
groups being made haphazardly
quota
(non-probability sampling)
sampling the same elements over & over again, based on past history or performance
purposive (judgmental)
(non-probability sampling)
the test or scale is measuring what it is supposed to be measuring
validity
each item on the scale subjectively appear to be related to the construct in question
face validity
the range of the construct & all its constructs & dimensions are adequately represented
it should cover the range of topics
content validity
taking a test score & comparing it against some criterion w/c is a more direct measure of that construct using correlation
criterion-related
the pattern of correlations w/ other measures make theoretical sense
construct validity
(theoretical validity)
expectations are shifted to the experimenters
“blind testing” experimenters & subjects don’t know what they are being tested for
experimenter bias
