9: Simple Linear Regression

Question 1

How many predictor variables are there in a simple linear regression?

Answer 1

1

Question 2

How do you calculate error variance for a simple linear regression?

Answer 2

sum(average-data point x)^2 - sum(predicted value of y - y data point)^2

Question 3

What does the error variance for simple linear regressions show?

Answer 3

the improvement in prediction using the regression model compared to the simplest model

Question 4

what is the non-parametric equivalent for a regression?

Answer 4

there is no non-parametric equivalent

Question 5

what does adjusted r^2 signify

Answer 5

r^2 is too optimistic to generalise to the general population, so adjusted r^2 is what we use

Question 6

what re the y= mx+c values in SPSS

Answer 6

Coefficients
m= Understandardized coefficients B 2nd value
c = Understandardized coefficients B 1st value
– there is beta value for m (standardised for SDs)

Question 7

what is the regression equation? and how do you use it for multiple variables

Answer 7

y=mx+c
y=mx+mx+c (you can find all the info you eed for this in SPSS)

Question 8

what does the t-value signify in regressions?

Answer 8

how much each individual predictor, separately, improves the prediction of y