Task 5: Setting up good research Flashcards
def: random sample
every subject has an equal opportunity to be chosen for the study
def: nonrandom/ probability sample
every member of population has known probability to be in sample
def: generalization
Possibility to apply finding from sample to whole population -> high degree of external validity
def: volunteer bias
participation in research has to always be voluntary BUT those who volunteer differ from those who don’t -> not representative for general population/ lowers external validity
internal validity
Trustworthiness of cause-effect relationship
def: representativeness of a sample
how closely matches characteristics of full population
def.: simple random sampling
just randomly picking a sample from a population
def.: stratified sampling (geschichtet)
divide in different groups, pick equal amount randomly out of every subgroup category
nearly same as multistage sampling during cluster sampling
def.: proportionate sampling
use percentages of subcategories to avoid overrepresentation
def: systematic sampling
chose every x^th subject, random start
def.: cluster sampling
divide population into subgroups (sometimes happens naturally)
choose random groups (=cluster)
def: sampling error
deviation between sampled results and results of entire population
def.: economic sample size
just enough to get valid result, not more
depends on amount of acceptable error and expected magnitude of population proportions
def: rival hypothesis (konkurrierend)
need for more research if several explanations produce same findings
def: confounding of variables
two variables occur together und their effects can’t be separated
eg split class, one improves skills faster than other, is just taugth at wrong time or differences in teacher, performance of the same teacher …
