CCP Stats Terms

Review for Exam 2

Mean Sentence

On average, the ‘‘context’’ is (blank)

Median Sentence

About 50% of the ‘‘context’’ is at or below (blank)

Interquartile Range Sentence

The middle half of ‘‘context’’ has a range of #

Standard Deviation Sentence

The typical distance of ‘‘context’’ from the mean is #

Undercoverage Bias

People aren’t included Ex. People at work, don’t have social media, etc

Non-Response Bias

People can’t answer/choose not to Ex. Some surveys people choose not to participate in

Response Bias

Responses aren’t truthful/accurate Ex. exercise, income, etc

Categorical data

Observations are placed in groups/categories

What makes a misleading graph?

1. Y-axis doesn’t start at 0

2. Bar graphs that use pictures instead of bars

Quantitative Data

Data dealing with numbers (makes sense to find the average)

Dot plot

Each dot represents one person/value

Shape (Distribution)

Skewed left, skewed right. Symmetrical

Center (Distribution)

The typical/ middle value

(use mean or median sentence)

Variability (Distribution)

How spread out the dots are

Low=not spread out

High=very spread out

Use IQR or SD sentence

Outlier

Values far from others/that don’t fit

What is included in a Histogram?

Bars touch

Include the left number!! Not the right

Can’t see every value

What is non-resistant to outliers and skew?

The mean and standard deviation

Factors of left skewed

Median

IQR

Mean < Median

Factors of symmetric

Mean

Mean = Median

Standard Deviation

Factors of right skewed

Median

IQR

Mean > Median

Marginal Probability

Only a single thing (B/C) Ex. What percent of the class considers themselves early birds?

Joint Probability

Two things together (A/C) Ex. What percent of the class sleeps best in complete silence AND considers themselves night owls?

Conditional Probability

A/B Ex. Of those who consider themselves early birds, what percent sleep best with background noise?

How to know if there is a correlation between variables?

Conditional percents the same- no/weak correlation

Conditional percents different- there is a correlation

If percents are within 5% of eachother, there is no correlation

Explanatory Variable

A variable that explains, influences, or predicts change in another variable

Response Variable

The outcome or result that changes in reaction to another variable

How to describe the relationship between two quantitative variables

D- Direction (+/-)

U- Unusual factors (outliers, clusters)

F- Form (linear/not)

S- Strenght (strong, weak, moderate)

Ex. There is a weak negative linear relationship between height and # of steps with no outliers or cluster of points

Correlation (r)

Closer to 1 or negative 1 = a strong correlation

Close to 0 = weak correlation

At 0 = no correlation

How to interpret the correlation

“the linear relationship between X and Y is strength + direction’’

Residual Formula

Actual-Predicted value

Prediction Equation

^y= b + ax (explanatory variable)

b = y-intercept

ax = the slope

How to choose a good model/residual plot?

Good: No pattern

Bad: Clear pattern

What is the linear and exponential prediction formula?

Linear: ^y = a + bx

Exponential: ^y = a x b^x

Equation of multiple regression models

Predicted (y) = Constant + (coef) variable + (coef) variable

What is a percentile and how do you find it?

The percent of values less than or equal to the context

Take the number in the scenerio over the total value

What is the formula for Z-score?

Value-mean/Standard Deviation

What is the Z-score sentence

‘‘The context is Z standard deviations above/below the mean.’’

What is the mean of Z-scores?

The mean of Z scores is always 0

What is the SD of Z-scores

The SD of Z scores is always 1