AP Statistics › How to do logarithmic transformations
What transformation should be done to the data set, with its residual shown below, to linearize the data?
take the log of the dependent variable
multiply the dependent variable by a constant k.
Add to the y-value of each data point
multiply the independent variable by
Nothing, the data set is already linear
Taking the log of a data set whose residual is nonrandom is effective in increasing the correleation coefficient and results in a more linear relationship.