Simple Regression 2

Question 1

What can the regression line be used to?

Answer 1

To estimate the magnitude of the effects of the predictor on the outcome.

Question 2

Where is SPSS is it possible to see the magnitude of the effects?

Answer 2

In the model summery.

Question 3

What is R in the model summery?

Answer 3

Simple correlation between the two variables

Question 4

What is R square in model summary?

Answer 4

Coefficient of determination. The proportion of variance in y explained by x.

Question 5

What is adjusted R square?

Answer 5

Adjusted measure of R square accounting for possible overestimation.
An estimate of R square in the population

Question 6

What is the standard deviation of the residuals in model summary? (Std. Error of the estimate)

Answer 6

The formula shows how it is calculated.

Question 7

How is R square calculated?

Answer 7

SSM/SST. It gives the portion of y which is explained by the variance of x. If two circles overlap it’s the overlapping area which is the explained variance.

Question 8

What has the same value in simple regression but not in multiple regression?

Answer 8

Beta has the same value as r in simple regression.

But beta and r is not the same in multiple regression.

Question 9

What does beta mean?

Answer 9

It means that for every standard deviation increases in variable x, the predicted score on the other variable increases by the number of beta standard deviations.

Question 10

What is beta (or standardized regression coefficient) about?

Answer 10

It refers to how much the value of the outcome increases or decreases, as the value of the predictor increases 1 SD unit.

Question 11

What is beta in a scatterplot?

Answer 11

It’s the slope of the regression line (which tells how many standard deviations the outcome will increase as the predictor increases 1 SD.

Question 12

What are beta useful?

Answer 12

It’s useful to compare the effects of the two variables.

Question 13

How do you report results of simple regression?

Answer 13

Beta symbol = .x, p = .x.

Question 14

Why are causality important to think of when doing regression?

Answer 14

Regression does not say anything about causality.

One variable does not cause an affect in the other variable. There is a relationship but no cause and effect.

Question 15

Why is correlation not an experiment?

Answer 15

Because we haven’t manipulated the independent variable.

Question 16

When it is possible to claim causation from correlational evidence?

Answer 16

• we can claim that x temporally precedes y. (E.g age)
• we are relatively confident that the found relationship between x and y is not due to the fact of a third variable determines x and y.
• the idea that x determines y makes good theoretical sense and it’s supported by independent evidence.