AP Stats 11-12

simulation

mimics reality by using an RNG to represent the outcome of real events

trial

each repetition or observation of an experiment

response variable

a trial's outcome

reason why we take samples

we cannot examine the entire population

sample size

less than 10% of the population

voluntary response bias

individuals choose if they can participate in the sample; always invalid

convenience sampling

sampling in the easiest way; not representative of the population

bad sampling frame

when samplers cannot reach some people so they are not sampled

undercoverage bias

a group of the population is either under or unrepresented in a sample

nonresponse bias

when a large fraction of those sampled fails to respond

response bias

anything in a survey design that influences responses

stratified sampling

some of all; implies a common characteristic

simple random sample

each combination of people has an equal chance at being selected

census

when one tries to sample the entire population

cluster sample

all of some; physically connected

systematic sample

obtained by selecting every nth individual from the population

multistage sample

combining multiple sampling methods

pilot

trial study

statistic

numerical summary of a sample

representative sample

accurately reflects corresponding parameters

standard error/deviation

typical deviation from the mean

correlation does not mean

causation

association does not mean

causation

observational study

researchers observe choices

retrospective study

studying historical data

prospective study

identify subjects in advance then observe as events unfold

experiment

imposes a treatment on a subject

4 necessities of an experiment

control, randomization, replication, blocking

factors

the thing that is changed in an experiment

control treatment

the baseline treatment applied to control group

blinding

method used so that units do not know which treatment they are getting

double blinding

subject and data recorder do not know the treatment

matching

subjects similar in ways not under study may be matched and then compared with each other on the variables of interest

confounding variable

a factor other than the independent variable that might produce an effect in an experiment

placebo

"fake treatment" helps for blinding