population
desired group to be understood
Simple Random Sample (SRS)
choosing a group from the population so that ever individual and every group of individuals if likely to be chosen
Stratified Random Sample
split the population into groups (strata) that are the same (homogeneous) and take an SRS/simple random sample from each strata
systemic sample
choose a random starting point and use equal intervals to move to every next individual until you have as many as you need
cluster sampling
split the groups into groups that are heterogeneous (different) then randomly select the groups themselves and sample everyone in those selected groups
undercoverage
when some members of a population cannot or are less likely to be included in a sample (calling landlines)
nonresponse
when an individual is part of a sample/survey but chooses not to respond or cannot be reached; different from voluntary because people cannot be put into a sample and not respond
response bias
patterns of inaccurate results (wording, interviewer, lying, etc.)
observational study
no treatment, can't show causation
experimental study
treatment imposed, shows causation
treatment
what is done/not done to the experimental units
control group
used to provide baseline data for comparison
double blind
researchers/subjects are unaware of what the treatment given
single blind
when only the subjects are unaware of the treatment being given
placebo effect
when a fake treatment works/gives wanted results
KEY 1: comparison
2 or more treatments
KEY 2: random assignment
taking samples and randomly placing in needed groups
KEY 3: control
keeps all other variables besides treatments constant
KEY 4: replication
using enough exp. units to distinguish differences
blocks
groups of experimental units which are similar
randomized block design
separate subjects into block and then randomly assign treatments within each block
matched pair design
this is a block design of pairs where subjects are randomly paired and each individually assigned treatments, and is repeated and given treatment in a random treatment
statistically significant
results are too unusual to have occurred purely by chance, must be < 5% to be so