AP Stats Chapter 5

systematically favoring certain outcomes

bias

2 types of studies

observational, experiments

observational studies: observe individuals and measure varialbes to infer up to ______

population

experiments: _______ on individuals and obsoerve their responses to infer _____

impose some treatment, causation

Biased sampling methods

• people who choose to respond (biased bc opinionated ppl tend to respond)

• sample ppl easy to reach (biased bc doesn’t reprsent pop)

voluntary response, convenience sampling

How do you use a Random Digit Table to create a sample?

Every group of size n has equal chance of being selected

SRS - simple random sample

Take an SRS from diff strata of population & combine into 1 sample *minimizes variability between samples

Stratified Random Sampling

divide cluster into grps that are representative of population. take an SRS of clusters and sample ALL individuals in those clusters

cluster random sample

randomly select starting # then select every nth individual

systematic random sample

types of bias

subjects voluntarily choose to be in the sample, and people usually volunteer only if they have strong opinions.

leaving parts of pop out of sample

individuals are chosen but don’t respond

respondents lie (illegal or unpopular question)

wording question in a leading way

Voluntary response bias

Undercoverage

Non-response

Response Bias

Question Wording Bias

individuals from pop we are choosing from

Sampling Frame

What can we conclude from an Observational Study?

When can use our sample to make inferences about the population?

Association or correlation, not cause and effect

When our sample is representative of population

a phenomenon where a person experiences improvement in their condition after taking an inert substance or receiving a fake treatment

placebo effect

What is the defining feature of an Experiment? Why is it important?

Impose treatments

We can finally infer cause and effect

4 principles of experimental design

also draw the outline of an experiment

What is required for replication?

Larger samples

Separate similar subjects into blocks, then randomly assign from each block

block design

What is the goal of Random Assignment?

What is the goal of Random selection?

in experiments, All variables split approx 50/50 → similar treatment grps

in observational studies, sample is representative of population so we can infer up to pop

If difference in response variables is larger than we’d expect based on random chance

Statistically Significant

Match similar subjects ahead of time randomly assign one to each treatment (eg. twins, left/right body part). It maximizes the similarity between treatment groups: more precise results. 

Matched Pairs design

To be blind in a study means the _____ don't know which treatment they are receiving, while the researchers do. (or particpants know but researchers don’t) To be double-blind means that neither the ______ nor the _____ know who is getting which treatment

participants, participants and researchers

Vocabulary for experiments

• What you are experimenting on

• the explanatory variables

• the choices you have for each factor

• the combinations you will test

Experimental units:

factors

levels

treatments

Random sampling _____ involves selecting an item from a population, recording it, and returning it before the next selection, allowing for the possibility of selecting the same item multiple times.

In contrast, Random sampling ______ means an item is not returned to the population after selection, so each item can only be selected once.

with replacement, without replacement

Lurking vs confounding variables

______: scatterplot correlation from common cause

______: experiment causation clouded by some other variable

lurking

confounding