Download the data set here:

https://drive.google.com/file/d/1CH0TFlSR4OKlPbqJrWiTCQkanSCtjI20/view?usp=sharing

You need Rstudio to do some of this problem.

Problem 2 A chemist studied the concentration of a solution over time. Fifteen identical solutions were prepared. The 15 solutions were randomljyr divided into five sets of three, and the solutions' concentration were measured after 1, 3, 5, 7, and 9 hours. Download the data le SolutionConcentrationcsv from Canvas. We want to model the relationship between time and cone . Fit the following models to these data. " J' 2 Bo + 3135 + E I ln[y]= 135+ [3136+ E " ag.) + 13: + E " 12Iy=igo+l31x+5 a. Create a scatterplot with time on the horizontal axis and cone on the vertical axis. Add the fitted regression line to the data. You need to do this for each of the four models since the y axis changes in each case. Which model fits the data the best based on the plot? b. Examine model diagnostics for the four models. Which model fits the data the best? Call this your final model. Is this the same model as Part a? c. How much of the variation in the concentration does your final model account for? d. Are any of the observations outliers, high leverage points, or inuential points in your nal model? Please use -2 and 2 as the thresholds for the standardized residuals because the sample size is small