Lecture 10

Question 1

Lecture 10:

Who developed the Bivariate Regression?

Answer 1

Sir Frances Galton

Question 2

Lecture 10:

What is Bivariate Regression?

Answer 2

A way to make predictions by looking at how independent variables impact dependent variables (using scores on 1 variable to predict scores on another)

Question 3

Lecture 10:

What is another way to say Bivariate Regression?
- & very simple definition

Answer 3

Also known as Simple Linear regression & estimates future outcomes from present ones

Question 4

Lecture 10:

How are Correlations used for Predictions?
- what does a larger correlation mean?

Answer 4

Correlations can be used as the basis for predicting X based on Y values
- larger correlation coefficient (r) = more accurate prediction (explained variance, r^2)

Question 5

Lecture 10:

What is the equation for Regression?
- explain what each letter means

Answer 5

Y’ = a +bX

• Y’ = predicted value of Y; not the actual Y value
• b = slope (direction 7 strength of relationship)
• a = intercept (point where regression line crosses Y axis or point where x = 0)

Question 6

Lecture 10:

What is the Error of Prediction?

Answer 6

The distance of the actual measure from the line of best fit (prediction)

Question 7

Lecture 10:

What is an error in prediction called?

Answer 7

Called the Residual (Y’)

Question 8

Lecture 10:

When determining error in your prediction, what is the Residual?

Answer 8

The vertical distance between actual data point and the line of best fit (best fit line represents best prediction of Y for any X value)

Question 9

Lecture 10:

What should the sum of the residuals be? - Sum(Y-Y’)

Answer 9

Sum(Y-Y’) = 0; all the Negative residuals will cancel with the positive residuals

Question 10

Lecture 10:

Review the calculations!!