Correlation, Linear Regression 2

Question 1

Why is extrapolation unreliable?

Answer 1

It goes outside the range of the data

Question 2

How do you know if the data is non-linear?

Answer 2

If many of the points clump together, it is non-linear therefore you can transform it to make it more linear e.g. taking the natural log of both sides (ln) if equation has e

Transforming the data also makes the variation about the line of best fit constant