Show:
Correlation
Regression
Effect Sizes
Formulas / Hypothesis Testing
100

What kind of data do you need to run a correlation?

Interval or ratio.

100

What kind of data do you need to run a regression?

Interval or ratio.

100

What is our effect size measure for a correlation?

The correlation itself (r).

100

Name this equation: y-hat = a + bx


Linear regression equation – the equation for a straight line.

200

What is Pearson’s r?

Correlational coefficient

• Measure of the linear association between variables (the extent the plotted data form a straight line).

200

Why would you use a regression instead of a correlation?

Varies.

⚠️ • Consider: Is there a difference between causation and correlation?

200

What is our effect size measure for a regression

Coefficient of determination (R2; R-squared).

200

How do you calculate degrees of freedom for a correlation?

df = n - 2

300

What is a positive correlation? What is a negative correlation?

As one variable increases, the other increases; as one variable increases, the other decreases.

300

In a regression, what do we call the IV? The DV?

X and Y

300

What are our effect size measures for correlations?

The correlation is an effect size measure. We also use the coefficient of determination.

300

What is b in the following regression equation: y-hat = a + bx

The slope of the line

  • The value of b is interpreted as: increase in y for every unit increase in x).
400

What two pieces of information does the correlation coefficient provide?

1.) Strength (i.e., magnitude) of the relationship 

2.) Direction of the relationship 

400

True or False: You can only have one predictor in a regression.

❌ False.

400

What is the coefficient of determination if r = .3?

It’s r-squared so it is...

• R2 = .09.

400

What is a in the following regression equation: 

y-hat = a + bx

The y-intercept.

  • Value when x = 0.
500

What does a correlation of 0 mean?

There is no LINEAR relationship between variables. (But there could still be a nonlinear relationship.)

500

Can you ever include a categorical variable in a regression? If so, how?

Dummy code

500

If the coefficient of determination (r-squared) is .4, we have a _____ effect.


Large.

Coefficient of Determination Thresholds (r-squared):

• Small = 0.01

• Medium = 0.09

• Large = 0.25

500

The formula for Pearson’s r is comprised of _____ and _____.


1.) Variances

2.) Covariances