AP Statistics - Chapter 4 Review

Why is a Simple Random Survey a good survey?

It is random

What are some bad sampling methods and why?

Personal choice. Convenience Survey and Voluntary Surveys

What are the forms of bias in sample surveys (give their definitions too)?

Undercoverage: When certain members of the population are unaccounted for or are less likely to be included in the sample.

Nonresponse: When an individual is picked as part of a sample but chooses not to be included.

Response Bias: Pattern of inaccurate responses due to the wording of a question.

What is the difference between Sample and Sampling?

Sample: Includes people/things you select from the population.

Sampling: Method of selecting samples.

What is a Simple Random Sample?

A good sampling method where every group is equally likely to be chosen. USE HAT METHOD

What is a population?

Entire group of individuals we want information about.

What is a convenience sample?

Sampling individuals that are easy to reach. BAD SAMPLE BECAUSE THEY INVOLVE PERSONAL CHOICE

What is a voluntary response sample?

People choose to be in the sample. BAD SAMPLING METHOD

What is a block?

Group of experimental units that are known to be similar.

What is a random block design?

Separate subjects into blocks then randomly assign to treatments within each block.

What is a matched pairs design?

Experiment to compare two treatments that uses blocks of size 2. Very similar experimental units are pairs then randomly assigned to a treatment. OR giving each subject the same treatment in a different order.

EX: Two highest students are grouped together and each one is separated into different treatments.

What is a “Modified” Matched Pairs?

Each subject received both treatments but in a different order.

What is the difference between Block and Stratified?

Block is meant for experiments
Stratified is sampling, how we pick people.

What is a parameter?

Number that describes some characteristic of a population.

What is a RANDOM SAMPLE?

Sampling method that involves using chance to determine which member of a population are included.

What is a census?

Collects data from every individual in the population.

What is a statistic?

Number that describes characteristics of a sample.

What is bias?

If something is very likely to under or overestimate a value in a statistical study.

What is an observational study?

A study that observes individuals and measures variables of interest but does not influence the responses.

What is confounding?

When two variables are similar in a way that their effects on a response cannot be distinguished from one another.

EX: Two very smart people

What is a treatment?

A specific condition applied to the individuals in an experiment.

What is a control group?

Group used to serve as a baseline for comparison. Treatment group that gets no treatment.

What is an experiment?

Measures an outcome of a study.

What is an experimental unit?

Where a treatment is randomly assigned to an experiment. Humans are called subjects.

EX: Waxed boxes and plastic boxes are experimental units that will be tested under different temperatures

What is a placebo effect?

Some subjects will respond favorably to any treatment.

what is a placebo?

A treatment has no active ingredient but is like other treatments.

What is a factor?

An explanatory variable that is manipulated and may cause a change in the response variable.

What is double-blind?

If neither the subjects or those who interact with them know which treatment a subject is getting.

How do you use a random number table in an SRS?

Label all the people in the population, look at the table and select the amount of people being sampled (EX: 10 people from a population of 100). From there, say that a random number from 001-100 will be chosen 10 times.

When you’re asked how to avoid bias what do you always say?

Take a simple random sample from the population.

What is a stratified random sample?

Splits population into groups (strata’s) and chooses an SRS from each group. Strata is based on some variable that is known or suspected to alter the results.

What is a cluster sample?

Splits population intro groups based on location (clusters). Sample includes every individual in each selected cluster. They are used for efficiency.

What is systematic sampling?

Selects members from a larger population from a list. A random starting point is chosen and then you select every n’th member for your study.

What is a control?

Overall effort to minimize variability and confounding between variables.

What does statistically significant mean?

When the observed differences in the results of an experiment are so large that it is unlikely they occurred by chance.

To make an inference about a population it has to be __________

random selection

To make an inference about cause-and-effect it has to be an experiment with _____

random assignment

Write out that box with the yes’s and no’s

Random Assignment W/ groups
Yes no
Yes Inference to Inference to population=yes

cause and effect cause and effect = no
yes

No Inference to
population=no

cause and effect cause and effect = no
yes

What is a major principle of good design for all experiements?

Comparison to a control, replication, randomization.

When can you make an inference about cause and effect?

When individuals were randomly assigned to groups.

If the probability is ABOVE 5% what do you say? BELOW 5?

This experiment is NOT statistically significant because the probability of _______ exceeds the 5% threshold. Therefore, this is likely to occur by chance.

This experiment is statistically significant because the probability of ____ is less than the 5% threshold. Therefore, these results are unlikely to occur by chance alone.

When can you make an inference about a population?

When you have random selection.

When can you ONLY use Statistically significant?

ONLY if it is an experiment, NOT an observational study.

When mentioning any sort of Bias, you need to mention that it will ______

overestimate/underestimate

What is the benefit to using a representative sample of the target population?

Using a representative sample from the target population can cause inferences to be make which can be relayed back to the population.

When using the hat method for a randomized experiment, be sure to say __________

without replacement

What does NOT statistically significant mean?

The results were no better than what could be expected by chance through random assignment of the treatments.

What is the advantage to using a randomized block design over a completely randomized design?

Usually, a random block design takes into account hidden variables.

What is blocking?

When you separate subjects into groups and then assign the treatment there to reduce confounding.