AP Statistics Chapter 13: Experiments

Explain the difference between an observational study and an experiment

In an observational study, researchers don't assign choices. Experiment, they randomly assign choices

Explain the differences between experimental units and subjects

Subjects- Humans

Experimental Units-More generic (other indvs. such as rats, days)

Why is it necessary to assign subjects to treatments at random

To be able to justify relationships

Describe the four Principles of Experimental Design

Control - Making conditions as similar as possible for all treatment groups

Randomization - Allows to equalize effects of unknown or uncontrollable sources of variation

Replication - Apply treatment to number of subjects

-Only with replication can estimate variability of responses

Blocking - Ability of randomizing to equalize variation across treatment groups

Explain what is meant by a control group

What groups are compared to. What group is receiving control treatment.

Define Statistically significant

big enough (significant enough) difference between results of data groups

What is the purpose of using a single-blind or double-blind experiment

To avoid bias

What is a placebo? What is meant by the placebo effect?

Placebo- A "fake" treatment that looks just like treatments being tested

Placebo effect- Highlights importance of effective blinding and importance of comparing treatments with a control

What is the purpose of using blocking in an experiment

Isolate variability attributable to differences between the blocks, so you can see the differences between blocks.

How might confounding affect the results of an experiment

Can't tell whether any effect we see was caused by our factor or confounding variable, or both working together.

observational study

researchers don't assign choices, simply observe them

retrospective study

researchers identify subjects and collect data

prospective study

identify subjects in advance and collecting data as events unfold

experiment

manipulate factor levels, randomly assign subjects to treatment levels, compare responses of groups

random assignment

needed for experiments to justify claims

subjects

humans being experimented on

participants

humans being experimented on

experimental units

generic term of other indvs. (rats, bacteria)

factors

explanatory variable to manipulate

levels

specific values that the experimenter chooses for a factor

treatment

combo of specific levels from all the factors that an experimental unit recieves

block

ability of randomizing to equalize variation across treatment groups work best in the long run

completely randomized experiment

subjects randomly assigned treatment

statistically significant

big difference between results of treatment groups

control

making conditions as similar as possible for all treatment groups

control group

what group is receiving control treatment (comparison)

single blind

when indvs. of one group that can effect outcome is blinded

double blind

when both groups are blinded

placebo

"fake" treatment that looks just like treatments being used

placebo effect

highlights importance of effective blinding and importance of comparing treatments with control

matching

compare treated and non treated in observational study

confounding

levels of 1 factor associated with levels of another factors