Sampling and Sample Data Flashcards
population
collection of objects with properties
elements
the objects in the population
cencus
a study of the whole population
often unfeasible due to numbers and limited resources
sample
selection of one or more elements of the population
statistical inference
understand the whole population by studying a sample from it
representative sample
properties of sample match unknown properties of population
non-representative sample
biased
when is bias almost certain
samples are chosen from people who volunteer
samples are drawn from only part of a population
samples are chosen for their convenience
samples are selected by someone with a vested interest in the outcome
how to avoid biased sampling
choose elements at random
choosing at random
each element of the population has exactly the same chance of being selected
what are the two ways of random sampling
with replacement and without replacement
when do the two types of random sampling essentially become equivalent
when samples are much smaller than the population
systematic sampling
elements selected according to a random starting point and a fixed periodic interval
stratified sampling
useful when the population splits into distinct groups where the groups are quite different
together the groups cover the whole population
cluster sampling
useful when the population splits into distinct clusters where the clusters are similar
take a random sample of clusters and then take the elements from within these clusters