Exponential & Power Models: Transforming Bivariate Data with Logs

When your scatterplot and residual plot suggests a non-linear relationship. You can transform the data using logarithms to achieve linearity. In this video we use the List function on a TI-84 Plus to first transform the response variable by taking its log (which would work if the true relationship is exponential) and then also transforming the explanatory variable by taking its log (which would work if the true relationship is a power model). We use the Stat Plot options to check the scatterplots and residual plots as well as the LinReg(ax+b) function. We also use our transformed equation to predict the response varibable for a given value of the explanatory variable.