AP Stats Unit 3 Review

Randomizing

Protects us from the influence of all the features of our population by making sure that the sample properly looks like the population

Simulation

mimics reality by using random numbers to represent the outcomes of real events

Response Variable

A trial's outcome

Census

Taking a sample of the entire population

Trial

Each time we obtain a simulated answer to our question

Population Parameter

A parameter (summary) of a population

Sample Statistic

A summary of a sample

Representative Sample

A sample that reflects the corresponding parameter accurately

Simple Random Sample

A sample where each combination of people has an equal chance of being selected

Sampling Frame

The list if individuals from which the sample is drawn from

Sampling Variability

The sample-to-sample differences from the use of randomness

Cluster Sampling

-Split into representative clusters and then taking a census of one whole cluster

-selected for reasons of efficiency, practicality, or cost

-Example: sampling neighborhoods and selecting one whole block

Multistage Sampling

-Sampling schemes that combine several methods

-Most multistage use some sort of stratification and cluster sampling

Systematic Sampling

-Selecting individuals systematically but still randomly

-Example: selecting the 10th person in line at a waterpark and then asking every 5th person after that

Voluntary Response Sample

A large group of people are invited to respond, and those who respond are counted

Stratified Random Sampling

-The population is grouped into similar homogeneous groups and then simple random sampling is used

-Benefit is reduced sampling variability

-Example: splitting men and women into two groups to test the effect of shampoo on hair volume

Convenience sampling

-Sampling the individuals who are convenient or us to sample

Observational Study

-Researchers do not assign treatments but rather they observe them

-treatments not imposed

-Example: Observing sleeping patterns of male/female cats

-CAN NOT draw cause+effect because randomization is not present

Retrospective Study

The subjects are selected and their conditions are determined.

-Gathering data on something that has already happened

Prospective Study

Subjects are followed to observe future outcomes. No treatment are deliberately applied.

-Subjects identified and then observe results

-Example: observing success of high school students in college

Experiment

-An experiment manipulates factor levels to create treatments, randomly assigns subjects to these treatments, and compares the responses

-Example: Tip % base on # of candies received

Factor

-A variable that is manipulated in the experiment

-Example: in an experiment comparing tip % for numbers of candies, the number of candies is the factor

Response

A variable whose value are compared across different treatments

Experimental Unit/Subjects

Individuals whom the experiment is performed on

Treatment

The different controlled processes applied to the experimental units

Experimental Principles

Control:

Randomization:

Replicate:

Blocking:

Statistically Significant

When an observed difference is too large for is to believe that it is likely to have occurred naturally

Control Group

A baseline treatment level that provides a basis for comparison

Blinding (Single/Double)

not knowing what treatment level is being applied to you

Placebo

-a treatment known to have no effect

-The Placebo effect is the tendency of humans to show a response when a placebo is administered

Blocking

-Gathering similar experimental unit groups together in order to rid of variability due to comparing different test subjects

-Example: Blocking men and women to see how they react to a certain medicine

Matching

Matching subjects of like qualities to reduce variability

-Example: picking teams in a pickup basketball game by putting the two best players on opposing teams

Confounding

When levels of a factors cannot be associated because they are too many variables unaccounted for

Lurking Variable

Underlying variable to addressed

Pilot

A trial run of a survey to a smaller group that you plan to eventually give to a larger group

Population

the entire group of individuals we hope to learn about

Sample

A representative subset of a population examined in hope of learning about the population

Sample Survey

A study that asks questions of a sample drawn from some population in hope of learning something about the entire population

Undercoverage Bias

Some portions of the population are not sampled at all or are given a smaller representation than deserving

-Example: Telephone surveys on weekdays exclude those working

Response Bias

-Refers to anything in the survey design that influences the responses

-Example: Questions worded in a biased manner

Nonresponse Bias

-Bias introduced when a large number of those sampled fail to respond

-Example: Telephone polls conducting during work days when many people can't pick up

Sample Size

The number of individuals in the sample

Voluntary Response Bias

Biased towards those with strong opinions and is not representative of the population as a whole