AP Stats. Unit 4: 'Designing Studies' Vocab

Convenient Sample

Sample is easy to reach

Simple Random Sample (SRS)

1) Label individuals

2) Randomize

3) Select

High Bias

Sample mean is consistently much greater (overestimate) or much lesser (underestimate) than the population mean

Low Bias

Sample mean is consistently close to the population mean

Population

All individuals (or "data points") of interest in a study

Sample

Smaller, manageable group of the population that is actually studied or measured

Voluntary Response

People choose to be a part of sample

Stratified Random Sample

1) Split population into groups (strata)

2) Choose an SRS from each strata

"Sample some from all groups"

Strata

Group that contains individuals with shared (homogeneous) attributes/characteristics

Homogeneous Grouping

Grouping individuals with shared attributes/characteristics

Heterogeneous Grouping

Grouping individuals with different attributes/characteristics

Good Estimate

Low bias, low variability

Cluster Sample

"Sample all from some groups"

System(atic) Random Sample

1) Choose a random starting point

2) Use equal intervals

Undercoverage (Bias)

Some people are less likely to be chosen

(Ex: calling landlines, surveying homeowners)

Nonresponse (Bias)

People can't be reached or refuse to answer

(Ex: don't answer, hang up on phone call)

Response Bias

Problems in the data gathering instrument or process

(Ex: people lie (self reported response), wording of question)

Confounding Variable

Influences the response variable

(Connected to the explanatory variable)

Factor

Another term for the 'explanatory variable'

Observational Study

No treatment imposed

Experiment

Treatments imposed

What kind of Study can produce Correlation?

Observational Study

What kind of Study can produce Causation?

Experiment

Experimental Unit

What/Who treatment was imposed on

Roughly Equivalent

Of groups produced by random assignment

Assumed to be relatively "balanced" b/c of this process

Unique

There are no others (repeats)

Treatment

What is done/not done to experimental units

(Combinations of)Levels of the explanatory variable(s)

Single Blind

When subjects don't know about treatments

Double Blind

When experimenters/researchers don't know about treatments

Placebo

A fake (inactive) treatment

Placebo Effect

When a fake treatment (placebo) works

Well-Designed Experiment

1) Comparison

2) Random assignment

3) Replication

4) Control

Random Assignment

1) Label

2) Randomize

3) Assign

Blinding

When either subjects or experimenters/researchers don't know about treatments

Retrospective Study

When investigators examine past data for a sample of individuals

Prospective Study

When investigators follow a sample of individuals into the future collecting

data.

Explanatory Variable

Variable that is used to explain or predict changes in the response variable

Response Variable

Variable that measures the outcome of a study or experiment

Census

Method of data collection that gathers information from every single member of an entire population

Low Variability

Sample mean is consistently close to the population mean

High Variability

Sample mean is inconsistently spread away from the population mean

Completely Randomized Design

All experimental units are assigned to treatment groups entirely by chance

(Randomized) Block Design

1) Separate subjects into blocks

2) Randomly assign treatments within each block

Block

Group of experimental units that are similar

Matched Pairs Design

1) Subjects are paired and then randomly assigned to a treatment

2) Each subject receives two treatments

*Order of the treatments must be randomized

Statistically Significant

When results of an experiment are unlikely (less than 5%) to happen purely by chance

Cause convincing evidence the treatment caused the difference

Sample Mean symbol

x̄ ('x-bar')

Population Mean symbol

μ ('mu')

Randomize

1) No repeats

2) Shuffle

Comparison

Compares two+ treatments

Replication

More than one (repeat) in each treatment group

Control Group

Keeps other variables constant

Random Assignment purpose

To help show causation

Pair

Block with a size of: 2

Random Sample purpose

To help show correlation

Establishing Causation

Concluding a treatment causes changes in the response variable

Establishing Correlation

Generalizing our conclusions to the population from which we sampled