AP Stats

random sample

using a chance process to determine which members of the population are included in the sample

Simple random sample (SRS)

a random sample of n size is chosen in such a way that every group of n individuals in the population has an equal chance to be selected as the sample

population

the entire group of individuals that we want info about

census

attempt to contact every individual in the population

sample

the actual part of the population examined to gather info

connection between sample and population

we look at sample to make conclusion about population

sampling method

refers to the process used to choose the sample from the population

bias

sampling method that systematically favors certain outcomes

stratified random sample

start by classifying the population into groups of similar individuals, called strata, then chose a SRS in each stratum and combine the SRSs to

cluster sampling

divides the population in clusters, some of these groups are randomly selected. then all individuals in the chosen group are selected for the sample. clusters should look like the population

convenience sampling

choosing individuals who are easy to reach, displays bias

voluntary response sample

consists of people who chose themselves by responding to a general appeal, displays bias

undercoverage

when members of the population cannot be chosen in a sample

nonresponse

when an individual chosen for the sample cannot be contacted or refuses to participate

response bias

a systematic pattern of inaccurate answers in a survey (do not answer to not be incriminated)

inference

why we study stats, take a sample to make conclusion about population

margin of error

how far we expect the sample to be from the population (does not correct for bias. only sampling variability)

benefit of increasing sample size

increases precision (not accuracy)

systematic random sample

select a value k (based on population and sample you want) and then select every k individual to be in the sample (can be good or bad)

observational study

observes individuals and measures variable of interest but does not attempt to influence the responses

experiment

deliberately imposes treatment on individuals to measure their responses

distinction between observational study and experiments

observational studies often fail because of confounding variables and well-designed experiments take steps to prevent confounding

confounding in an experiment

occurs when two variables are associated and their effects on a response variable cannot be distinguished

treatment in an experiment

a specific condition applied to the individuals in an experiment

experimental units (subjects) in an experiment

the smallest collection of individuals to which treatments are applies. humans are called subjects

factors in an experiment

a value of the explanatory variable that treatments are made of

levels in an experiment

the different values of a factor

comparison in an experiment

use a design that compares two or more treatments

control group in an experiment

the group that does not receive active treatment or receives an existing baseline treatment

random assignment in an experiment

experimental units are assigned to treatment using a chance process

• solution to bias

• creates roughly equivalent groups of experimental units

control in an experiment

keep other variable that might affect the response the same for all groups

replication in an experiment

use enough experimental units in each group so that any differences in the effects of the treatments can be distinguished from chance differences between the groups. In a completely randomized design, the treatments are assigned to all experimental units completely by choice.

placebo

inactive treatment

placebo effect

thinking that someone got results from an inactive treatment

blind in an experiment

experimenter knows the treatment, but the subject does not know

double blind in an experiment

experimenter and subject do not know the treatments

statistically significant

observed effect is so large that it is unlikely to happen by chance

blocking

an experimental design technique where similar experimental units are grouped into blocks based on a shared characteristic. treatments are randomly assigned within each block.

matched pair design

common form of blocking for comparing 2 treatments. each subject receives both treatments in a random order. (sometimes similar subjects instead of the same person)

law of large numbers

states that if we observe more and more values of a chance process, the proportion of times that a specific outcome occurs approached a single number. this value is known as the probability of the event which we’ll write as a decimal

simulation

a way to imitate a chance process based on a model that reflect the situation