Probability Rules
Language of Data
Correlation & Regression
Experimental Design
Sampling and Bias
200

This type of event describes outcomes that have no common elements and therefore cannot happen at the same time

Mutually Exclusive Events

200

The complete set of all possible outcomes in a statistical experiment, often denoted by S

Sample Space

200

The critical statistical principle summarized as: Correlation does NOT imply this

causation

200

This principle involves treating all groups similarly except for the factor being tested, aiming to reduce variability from lurking variables

control

200

In statistical terms, this refers to a result that heavily favors an outcome in a manner that is unfair or subjective, often coming from non-random sampling methods

bias
400

For any two events A and B, this general rule is calculated as P(A)+P(B)−P(A and B)

General Addition Rule

400

The mean calculated from a sample, which is typically denoted by the notation xˉ (x bar)

sample mean

400

If r=−0.90, the value 0.81 represents what?

coefficient of determination (r^2)

400

This principle ensures multiple individuals are placed in each treatment level to estimate the variability of responses and increase the reliability of results

replication

400

The type of bias introduced when conducting a survey using only landline telephone numbers from a phonebook, as it systematically excludes segments of the population

undercoverage bias

600

What is the probability of A given A?

1, 100%
600

In regression, this variable is plotted on the y-axis, is what you hope to predict, and is also called the dependent variable

response variable

600

Defined as the vertical distance from an observed data point (y) to the regression line (y^), representing the prediction error

residual
600

The simplest experimental layout, suitable for homogeneous units, where treatments are assigned to individuals entirely at random

Completely Randomized Design (CRD)

600

This type of bias occurs when individuals selected for a sample cannot or do not provide responses, meaning the views of nonrespondents are missed

nonresponse bias

800

To correctly apply the Total Probability Rule, the preceding conditions (B events) should _______ the sample space? 

cover/completely partition
800

In an experiment, these are the specific conditions applied to subjects, representing a combination of specific levels from all factors

treatments

800

The type of variable that explains the spurious positive association between ice cream sales and shark attacks

lurking variable

800

This principle is used when experimental units are heterogeneous, requiring the grouping of individuals by a common characteristic before randomization

Blocking

800

A non-random sampling technique where subjects are selected based on the researcher's specialized knowledge or judgment, useful when targeting experts

purposive sampling/judgmental sampling

1000

DOUBLE OR NOTHING

If P(A) is .3, P(B) is .8, and P(B or A) is .7, what is P(A given B)?

.5, 50%, 1/2

1000

This numerical measure describes the entire population and is usually unknown, requiring estimation using a sample measure

parameter

1000

The preferred appearance of a residual plot, showing points scattered randomly around this central value, which validates the appropriateness of a linear model

scattered randomly and evenly around the x-axis

1000

The primary purpose of this design principle is to equalize the effects of unknown or uncontrollable sources of variation across treatments

randomization

1000

The process of obtaining samples from a population where every individual is given an equal chance of selection, eliminating human bias and ensuring a representative sample

random sampling

M
e
n
u