AP Stats Unit 2

Blaghh

What is a simple random sample? (SRS)

A method of selecting a subset of individuals from a larger population, where each individual has an equal chance of being chosen.

How to conduct a simple random sample? (SRS)

1. Label individuals (assign numbers/write names on slips of paper)

2. Use a random number generator or draw slips to select individuals. ( No repeats)

3. Select desired # of samples

What does a simple random sample (SRS) do?

Low bias, can represent a population and be applied to said population

What is the difference between volunteer bias and non-response bias?

Volunteer bias has no set number of participants and non response had a set number

Convenience sample

Researchers choose the easiest people to reach, high bias

Voluntary response

People volunteer, high bias

Stratified random sample

out of each group you do simple random sample, basically sample from a sample, homogeneous (even) subgroups are used SAMPLE SOME FROM ALL GROUPS

Strata def

Non-overlapping subgroups of a population divided by similar attributes or characteristics (selecting a whole row or column AKA homogeneous grouping)

A sampling method produces the best estimates if there is:

Low bias, low variability

Cluster random sample

Groups here are heterogeneous (various) random groups selected and all the entire group is selected not parts inside it SAMPLE ALL FROM SOME GROUPS (can be just one group out of multiple ex: only seniors)

Homogeneous groups

A group where all participants share similar or identical traits

Heterogenous groups

A group where groups are made of different individuals, making various mini-versions of the population, or trying to

systematic random sample

choose random starting point, use equal intervals

under coverage bias

Not everyone was surveyed who should have been surveyed

non response bias

Not all of the participants responded

response bias

Participants responses might have been affected by the conditions of the question where participants will provide inaccurate answers to a survey or in an interview

Observational

No treatment

Experiment

Treatment is imposed, allows to prove causation

Explanatory variable

Explains the reason for the results of the data

Response variable

The variable that comes out of treatment

Confounding variables

Outside factor that influences the results of a study, preventing causation

Random assignment doesn’t get rid of confounding variables, but..

makes the likelihood of them roughly equal between the two groups and indicates causation

Design an experiment format

x random ____, randomly assign- group 1 & group 2 - group 1 does no treatment/treatment & group 2 does treatment- compare data

Imposed variables

A factor or condition that is introduced, controlled, or manipulated by the researcher to observe its effects

Experimental units

what/who treatment- is being imposed on (ppl, plants, ect.)

Treatment

What is being done (or not done) to the experimental units

Replication

Assigning a treatment to more than one experimental unit (or participant) in each group, reduces random off-chance results

Random assignment will show..

CAUSE AND EFFECT (CAUSATION)

Well designed experiment has what four factors

1. comparison of 2 or more treatments

2. Random assignment

3. Replication (more than one instance in each treatment group)

4. Control keep all other variables constant

Random assignment steps

1 label

2 randomize

3 assign

For experiments, what is the equivalent of simple random sample for observation

Random assignment AKA Completely randomized design

For experiments, what is the equivalent of stratification for observation

Block design

Block design purpose

To prevent what could happen with random assignment which is putting one type of subject all into one group

Block design

Like randomized design but starting with two SIMILAR/SPECIFIC groups and randomly assigning those two groups to different treatments such as 1a 1b 2a 2b ect. ALWAYS compare individually and then together at the end

Matched pair design (for EXPERIMENTS)

Like block design but instead of two blocks there’s a lot of groups that you pair by two or more, for ex top two students and next two students and then flip a coin to see who gets what treatment

Randomized block design

Does not start with specific groups but randomizes them

Statistically significant

Off enough from the average to indicate whether the individual experiment’s results were an indicator that there was significance to the data

Statistically significant does or does not equal causation

does NOTTTTTT

What do you compare the 5% to

To the initial result and anything more extreme of it (count how many) over the total number of times repeated

The initial result and anything more extreme of it over the total number of times repeated (the 42%) indicates what?

How likely it ISSSS to happen JUST BY CHANCE

What does SRS stand for?

Simple random sample

What is a table of random digits (D Table)

idek