AP Statistics Unit 3 Notes
6 Steps of a Stimulation
Select components, when will a trial finish, determine response variable, run trials using RNG, summarize results, state your conclusion
1st step of simulation
select components
2nd step of simulation
when will a trial finish
3rd step of simulation
determine response variable
4th step of simulation
run trials using RNG
5th step of simulation
summarize results
6th step of simulation
state your conclusion
Parameters
population
Statistics
Sample
n
sample size
p
parameter proportion
p hat
statistic proportion
mu
parameter mean
x bar
statistic mean
Stratified sampling
externally heterogenous, internally homogenous
Cluster sampling
internally heterogeneous, externally homogenous
Four principles of experimental design
control, randomized, replicable, blocking
Control
control extraneous variables, not same as control group
Randomized
best chance of equalizing effects of unknown sources of variance, does not eliminate unknown variables
Replicable
Should be able to repeat with same sample or different sample, not have ridiculous amount of variability, one trial not enough
Blocking
Split trial up based on another variable you think might play a role (like stratify) before assigning treatments, break up each block into treatment groups, block is experimental design vs strata is sample design
Confounding variables
believe that x influences y, possible another variable influences y as well, can't tell which is causing y to happen combat by blinding, placebo, randomization
Lurking (hidden) 3rd variable vs confounding variable
lurking (correlation) is hidden 3rd variable related to x and y, makes it look like x and y are related but they aren't, confounding (experiment) is causation found between x and y is muddle by other variable influencing y