apstats 4.2 Flashcards
observational study
observes individuals and measures variables of interest but does not attempt to influence the response
experiment
statistical study where we do something to people, animals, or objects to observe the response
explanatory variable
variable that may help, explain, predict, or cause changes in the response
response variable
variable affected by the explanatory variable
treatment
condition applied to variables
factors
several explanatory variables
experimental units
smallest collection of individuals to which treatments are applied
subjects
human beings are called this
cause and effect relationship
experiments are the only way we can have convincing evidence for this, an observational study cannot establish a cause and effect relationship
confounding
when two variables are associated so that their effects on a response variable cannot be distinguished from each other
when identifying a confounding variable
explain how the variable is associated with the explanatory variable, explain how the variable affects the response variable
good experiment has these 4 components
comparison, random assignment, control, replication
comparison
uses two or more treatments to compare response variables
random assignment
experimental units are randomly assigned to treatments (method must be random using chance process, described specifically so people can recreate exactly)
control
efforts made to keep other variables that might affect the response variable the same for all groups
replication
more than one experimental unit in each group to tell the difference between effects of treatment and chance differences
completely randomized design
experimental units are assigned to the treatments completely by chance
control group
usually receives no treatment or an inactive treatment
placebo
dummy pill that contains no active ingredients is used for a control group in dug trials
placebo effect
subjects experiment improvement from a placebo
blinding
do not tell subjects what they are receiving
double-blinding
doctors and subjects are both unaware who is getting what
statistically significant
an observed effect is so large that it would rarely occur by chance
block
a group of experimental units in some way that is expected to affect the response to treatments
randomized block design
separates the experimental units into blocks and then randomly assigns experimental units to treatments
matched pairs design
matching pairs of similar experimental units into block sizes of 2
