Regression

Question 1

Variable types

Answer 1

Independent/explanatory- this is the one we control
Dependent/response- this is the one we measure

Question 2

Residuals

Answer 2

It is the vertical distance of a datapoint to the line of best fit
Residual= Observed- expected

Question 3

Least squares regression lie

Answer 3

y=a+bx
A line of best fit such that-
Sum of residuals= 0
Sum of squares of residuals is as small as possible
Menu 6-2 OPTN 3

Question 4

Interpretation of a and b

Answer 4

a- expected value of y when x is 0
b- expected change in y for every 1x

Question 5

Outliers and Anomalies

Answer 5

An outlier is any data pout with a residual of more than 3 s.d of y.
Do not remove outliers unless they are anomalies which are proven to not belong

Question 6

How to predict it

Answer 6

Either sub x into the equation
Or, use ŷ (OPTN down 4 5)

Question 7

Reliability

Answer 7

Interpolation- predict within the data (reliable)
Extrapolation- predict outside data
(not reliable)
Small residuals- reliable
Large residuals- not reliable