sampling Flashcards
(21 cards)
what is sampling
the process of selecting a smaller groups from a larger group to study the characteristics of the larger group
sampling used to
make predictions or generalisations about the whole population
population
larger group
sample
subset of the population selected to represent the whole group
representativeness
sample should accurately represent the population it cannot be biased
random selection
each member should have and equal chance of being selected
sufficient sample size
larger samples reduce random error making estimates more precise
central limit theorem
sample size of 30 is often sufficient for the sampling distribution of the mean to approximate normality
simple random sampling
number assigned to each member in sampling frame and number randomly generated
characteristics of simple random sampling
easy to implement and provide unbiased estimate but not feasible for large populations and chance of sampling error
stratified sampling
population divided into groups based on characteristics and number from each group is selected randomly
characteristics of stratified sampling
accurate but requires knowledge about the population and may not be feasible if population has many strata
systematic random sampling
population is selected at regular intervals
characteristics of systematic random sampling
easy to implement, more efficient and useful when population is already ordered but chance of bias and not appropriate if population doesn’t follow natural order
cluster sampling
sample of clusters selected randomly, and survey everyone in selected clusters or take sample from each cluster
characteristics of cluster sampling
cost effective and useful when difficult to obtain a list of individuals but less precision and requires larger sample sizes
sampling error
difference between sample mean and population mean due to random variation in the sampling process
sampling error occurs
when studying a subset of the population
standard error
measures expected variation due to sampling error
standard error is crucial
for hypothesis testing to calculate confidence intervals and test statistics