AP Stats - chapter 4

census

collects data from every individual in the pop.

simple random sample (SRS)

good sample design - chosen in such a way that every possible combo of individuals has an equal chance of selection.

1. assign distinct number

2. select unique numbers

3. individuals w/ labels matching the chosen numbers from the sample

stratified RA

good sample design

1. classify individuals based on similarities that might affect the response

2. choose SRS from each group

3. combine SRS’s to form sample

cluster RA

good sample design

1. classify the pop. into groups of individuals (often formed by location)

2. choose an SRS of clusters

3. every individual from each cluster is added from the sample

convenience sample

poor sample design - choosing individual from a pop. that are easiest to contact

voluntary response sample

poor sample design - people essentially choosing themselves by responding to some general invitation. ex. survey

N

sample size

Bias

difference between facts and results

inference

the purpose of a sample is to give us information about a larger pop.

observational study

observed individuals and measures variables but does not attempt to influence response. cannot calculate causation, no treatments imposed.

confounding

occurs when 2 or more variables are associated in such a way that their effects on a response variable cannot be distinguished from one another.

experiment

deliberately imposes on individuals to measure responses. can prove causation/cause + effect relationships.

treatment

specific condition applied to individuals in the experiment,

factor

explanatory and response variable

explanatory variable

the variable that is manipulated or changed in an experiment to see its effect on the response variable

response variable

how we measured the success of the experiment

four principles of exp. design

1. comparison - compare 2 or more treatments

2. random assignment RA - eliminate bias

3. control - keep other variables that might affect the response equally

4. replication - use enough subjects so any difference in response can be distinguished from diffs in groups

placebo (effect)

dummy treatment that makes subjects feel as though they are receiving an actual treatment.

completely randomized experiment design

sample → groups → treatments → comparison

experimental logic

the logic behind an experiment depends on our ability to treat all subjects the same except for the treatments they recieve 

single blind

when the subjects don’t know which treatment they receive but the expirementors do

double blind

when neither the subject or the ones administering the treatments know which treatment is being administered

blocking

group of experimental units that are known beforehand to be similar in ways that are expected to affect their response to treatments. the assignment of subjects to treatments is carried out separately within each block. 

block design

population → block → RA → group → RA → treatment → compare groups → compare blocks

matched pair

type of block design where it compares two treatments with either the subject and themselves or the subject and someone extremely similar.